1a NION
1a NION
1a NION
PART I: GRAMMAR
A NEW
INTRODUCTION TO
OLD NORSE
PART I
GRAMMAR
THIRD EDITION
BY
MICHAEL BARNES
ISBN: 978-0-903521-74-1
Michael P. Barnes
University College London
July 1999
April 2004
Preface to third edition
The book has been corrected and revised throughout, and a postscript
added (pp. 262–3).
May 2007
Contents
Abbreviations and Symbols ........................................................... xii
1. Introduction
1.1 The aim of the Grammar .................................................. 1
1.2 What is Old Norse? ........................................................... 1
1.3 Old Norse and modern English ......................................... 2
1.4 Pronunciation .................................................................... 3
1.5 Orthography ...................................................................... 4
1.6 General advice to the student ............................................ 5
2. Pronunciation and Orthography
2.1 Old Norse .......................................................................... 8
2.1.1 Pure vowels .......................................................... 8
2.1.1 Exercise ................................................................ 9
2.1.2 Diphthongs ........................................................... 9
2.1.2 Exercise .............................................................. 10
2.1.3 Consonants ......................................................... 10
2.1.3 Exercise .............................................................. 12
2.1.4 Syllables ............................................................. 13
2.1.4 Exercise .............................................................. 13
2.2 Modern Icelandic ............................................................ 14
2.2.1 Vowels ................................................................ 14
2.2.1 Exercise .............................................................. 16
2.2.2 Consonants ......................................................... 16
2.2.2 Exercise .............................................................. 20
2.2.3 Syllables ............................................................. 20
2.2.4 The epenthetic vowel ......................................... 20
2.2.3/2.2.4 Exercise ......................................................... 21
3. Morphology and Syntax
3.1 Noun inflexions and their function ................................. 22
3.1.1 Number ............................................................... 22
3.1.2 Case .................................................................... 22
3.1.3 Gender ................................................................ 27
3.1.1/3.1.2/3.1.3 Exercise ................................................ 28
3.1.4 Basic noun inflexions ......................................... 28
3.1.4 Exercise .............................................................. 31
3.1.5 Examples of noun usage ..................................... 31
3.1.5 Exercise .............................................................. 37
3.1.6 Difficulties in recognising noun inflexions
and ways of overcoming them ..................... 37
3.1.6 Exercise .............................................................. 39
3.1.7 Important variations in noun inflexion ............... 39
3.1.7.1 Labial mutation ............................................ 39
3.1.7.1 Exercise ....................................................... 41
3.1.7.2 Front mutation ............................................. 41
3.1.7.2 Exercise ....................................................... 44
3.1.7.3 Breaking ...................................................... 44
3.1.7.4 Deviations from the basic endings .............. 45
3.1.7.5 Minor irregularities ...................................... 45
3.1.7.3/3.1.7.4/3.1.7.5 Exercise ................................. 46
3.1.8 Examples of noun inflexion ............................... 47
3.1.8 Exercise .............................................................. 53
3.1.9 The suffixed definite article ............................... 56
3.1.9 Exercise .............................................................. 58
3.2 Pronoun inflexions and their function ............................. 60
3.2.1 Personal pronouns: form .................................... 61
3.2.2 Demonstrative pronouns: form .......................... 63
3.2.3 Indefinite pronouns: form .................................. 65
3.2.4 Negative pronouns: form .................................... 66
3.2.5 Interrogative and distributive pronouns: form ... 67
3.2.1/3.2.2/3.2.3/3.2.4/3.2.5 Exercise .............................. 68
3.2.6 Examples of pronoun usage ............................... 68
3.2.6 Exercise .............................................................. 76
3.3 Adjective inflexions and their function ........................... 77
3.3.1 Number, case and gender ................................... 77
3.3.2 Definiteness ........................................................ 78
3.3.3 Degree (comparison) .......................................... 79
3.3.1/3.3.2/3.3.3 Exercise ................................................ 79
3.3.4 Basic adjective inflexions .................................. 80
3.3.4 Exercise .............................................................. 84
3.3.5 The free-standing definite article ....................... 84
3.3.5 Exercise .............................................................. 86
3.3.6 Examples of adjective usage .............................. 87
3.3.6 Exercise .............................................................. 95
3.3.7 Difficulties in recognising adjective inflexions
and ways of overcoming them ..................... 96
3.3.7 Exercise .............................................................. 97
3.3.8 Important variations in adjective inflexion ........ 98
3.3.8.1 Labial mutation ............................................ 98
3.3.8.1 Exercise ....................................................... 99
3.3.8.2 Front mutation ............................................. 99
3.3.8.3 Suppletive forms ........................................ 100
3.3.8.4 Deviations from the basic endings ............ 101
3.3.8.5 Minor irregularities .................................... 102
3.3.8.2/3.3.8.3/3.3.8.4/3.3.8.5 Exercise ................... 103
3.3.9 Examples of adjective inflexion ....................... 104
3.3.9 Exercise ............................................................ 111
3.4 Numerals ....................................................................... 115
3.4.1 The numerals and their inflexions .................... 115
3.4.1 Exercise ............................................................ 119
3.4.2 Examples of numeral usage ............................. 120
3.4.2 Exercise ............................................................ 123
3.5 Adverbs ......................................................................... 124
3.5.1 Adverb formation ............................................. 124
3.5.2 Inflexion for degree .......................................... 125
3.5.3 Examples of adverb usage ................................ 127
3.5.1/3.5.2/3.5.3 Exercise .............................................. 129
3.5.4 Adverbs and adverbials .................................... 130
3.6 Verb inflexions and their function ................................. 131
3.6.1 Person and number ........................................... 131
3.6.2 Tense ................................................................ 132
3.6.3 Mood ................................................................ 134
3.6.4 Voice ................................................................. 135
3.6.1/3.6.2/3.6.3/3.6.4 Exercise ..................................... 136
3.6.5 Basic verb inflexions ........................................ 137
3.6.5.1 Endings ...................................................... 137
3.6.5.1 Exercise ..................................................... 139
3.6.5.2 Vowel alternations ..................................... 140
3.6.5.2 Exercise ..................................................... 143
3.6.5.3 The -sk form .............................................. 144
3.6.5.3 Exercise ..................................................... 146
3.6.6 Finite and non-finite forms; principal parts ..... 146
3.6.6 Exercise ............................................................ 152
3.6.7 Preterite presents and other irregular verbs ...... 152
3.6.7 Exercise ............................................................ 155
3.6.8 Examples of verb usage ................................... 155
3.6.8 Exercise ............................................................ 164
3.6.9 Important variations in verb inflexion .............. 164
3.6.9.1 Phonological variation ............................... 165
3.6.9.2 Morphological variation ............................ 168
3.6.9.3 Idiosyncratic variation ............................... 169
3.6.9.1/3.6.9.2/3.6.9.3 Exercise ............................... 171
3.6.10 Examples of verb inflexion .............................. 171
3.6.10 Exercise ............................................................ 177
3.7 Prepositions ................................................................... 181
3.7.1 Prepositions triggering the accusative .............. 182
3.7.2 Prepositions triggering the genitive ................. 184
3.7.3 Prepositions triggering the dative ..................... 185
3.7.4 Prepositions triggering the accusative
and dative ................................................... 189
3.7.5 Prepositions triggering the accusative and
genitive ...................................................... 196
3.7.6 Preposition triggering the accusative,
genitive and dative ..................................... 197
3.7.7 Residual remarks .............................................. 197
3.7 Exercise ......................................................................... 199
3.8 Conjunctions ................................................................. 200
3.8.1 Coordinating conjunctions ............................... 202
3.8.2 Subordinating conjunctions .............................. 204
3.8.2.1 The particle er ............................................ 204
3.8.2.2 The particle at ............................................ 212
3.8.2.3 Interrogative pronouns and adverbs .......... 215
3.8.2.4 Other adverbial sentence introducers ....... 218
3.8 Exercise ......................................................................... 221
3.9 Residual points of syntax .............................................. 223
3.9.1 Sentence word-order ........................................ 223
3.9.1 Exercise ............................................................ 228
3.9.2 Word-order in noun phrases ............................. 228
3.9.2 Exercise ............................................................ 230
3.9.3 Impersonal constructions ................................. 230
3.9.3 Exercise ............................................................ 236
3.9.4 Accusative and infinitive .................................. 236
3.9.4 Exercise ............................................................ 240
3.9.5 Omissions ......................................................... 240
3.9.5.1 Objects ....................................................... 241
3.9.5.2 vera ............................................................ 241
3.9.5.3 Verbs of motion ......................................... 243
3.9.5 Exercise ............................................................ 243
3.9.6 Points of nominal syntax .................................. 244
3.9.6.1 Idiomatic uses of personal pronouns and
possessive adjectives .......................... 244
3.9.6.2 The genitive and dative of respect ............. 246
3.9.6 Exercise ............................................................ 248
3.9.7 Points of verbal syntax ..................................... 248
3.9.7.1 The perfect and past perfect ...................... 248
3.9.7.2 The passive ................................................ 251
3.9.7.3 The ‘dative absolute’ ................................. 253
3.9.7.4 Present participles expressing potentiality
or obligation ........................................ 254
3.9.7 Exercise ............................................................ 255
3.9.8 Points of syntax affecting more than one
type of phrase ...................................... 256
3.9.8.1 Adjectival and adverbial complements ..... 256
3.9.8.2 Agreement between subject, verb and
subject complement ........................... 257
3.9.8.3 -sk verb forms and ‘preposition adverbs’ .... 259
3.9.8 Exercise ............................................................ 260
3.9.9 Adverbial ok ..................................................... 261
A postscript on ‘impersonal constructions’ ................................. 262
References to linguistic terms explained in the Grammar .......... 264
Select glossary of linguistic terms not explained in the Grammar ... 266
Bibliography ................................................................................ 270
Abbreviations and Symbols
acc. accusative
act. active
adj. adjective
art. article
aux. auxiliary
comp. comparative
COMP complementiser
dat. dative
def. definite
f. feminine
gen. genitive
imp. imperative
indic. indicative
inf. infinitive
interrog. interrogative
m. masculine
n. neuter
NION I–III A New Introduction to Old Norse I: Grammar;
II: Reader; III: Glossary and Index of Names
nom. nominative
NP noun phrase
ON Old Norse
pass. passive
pl. plural
pos. positive
pp. past participle
pres. present
refl. reflexive
REFL. POSS. reflexive possessive
sg. singular
subj. subject; subjunctive
sup. superlative
vb. verb
* reconstructed form; ungrammatical sentence
> develops to
< develops from
~ zero ending
1 Introduction
From the point of view of the student, many existing grammars of Old
Norse suffer from two major defects. First, they are largely constructed
on historical principles and thus contain detail about earlier stages of
the language and linguistic development, little of which is of direct
use to someone seeking to acquire a reading knowledge of Old Norse.
Second, they assume a level of linguistic sophistication which the
school system no longer cultivates, and so leave unexplained many
things which to the modern student are opaque.
The present Grammar has been written with one aim only: to facili-
tate the learning of Old Norse for as wide a range of students as possible.
It therefore eschews historical digressions except where they throw
essential light on the workings of the language, and an attempt is made
to explain all but the most basic ideas, concepts and terms on their
first appearance or, failing that, in the Select Glossary (pp. 264–6).
The emphasis throughout is pedagogical, and the work thus represents
not so much a re-think of Old Norse grammar as a re-think of the ways in
which the basics of Old Norse may be best presented to the learner.
The term ‘Old Norse’ has been used in various ways. For some it is a
broad concept covering the language of Denmark, Norway and Swe-
den, as well as Iceland and the other Scandinavian colonies, through-
out the Viking Age (c. 750–1050) and the early and high Middle Ages
(c. 1050–1350). At the other extreme it has been taken to mean only
the Old Norwegian of the early and high Middle Ages. In the present
context it is used principally to signify the language of Norway in the
period c. 750–1350 (after which Norwegian changes considerably) and
of Iceland from the settlement (c. 870) to the Reformation (c. 1550 —
a date that sets a cultural rather than a linguistic boundary). Known in
modern Icelandic as norræna, in Norwegian as norrønt and in English
sometimes as Old West Norse, this type of speech is a western variety
2 A New Introduction to Old Norse
A major difference between Old Norse and modern English is that Old
Norse is a much more highly inflected language. Modern English still
has certain inflexions, by which is meant that words change their form
according to their function in a sentence (e.g. she came, I saw her; sg.
cat, pl. cats; pres. run, past ran), but Old Norse has a far greater number.
In English the function of a word can often be deduced from its posi-
tion in relation to other words. We understand:
Introduction 3
to mean that Olav was the one who saw and the old woman the one
who was seen because Olav precedes saw. Reverse the order and the
opposite applies. In a corresponding Old Norse sentence it would be
perfectly possible for the order to be reversed without a change in
meaning. Everything would depend on the inflexions. Thus:
and
both mean ‘Óláfr saw the old woman’, because the forms Óláfr and
konu flá ina gƒmlu are unchanged. If we wish the sentence to mean
‘the old woman saw Óláfr’, we must alter the forms of the words so that
Óláfr becomes Óláf and konu flá ina gƒmlu becomes kona sú in gamla.
It is obvious, therefore, that from the very start the student will
have to pay the closest attention to inflexions. Failure to do so will
result in the regular misunderstanding of Old Norse texts.
1.4 Pronunciation
1.5 Orthography
The scribes who wrote Old Norse did not conform to standardised
rules of spelling, any more than their counterparts in medieval Eng-
land and elsewhere. They wrote words more or less as they had been
trained to do at the scriptorium where they studied, although they might
also be influenced by forms in an exemplar from which they were
copying — and occasionally by their own pronunciation. The result is
that most Old Norse words appear in manuscripts in a variety of spell-
ings. In order to facilitate the making of grammars, dictionaries and
text books, therefore, and to help the learner, modern scholars have
adopted a normalised orthography for Old Norse. Some editions of
Old Norse writings, designed more for the philologist and linguist
than the literary reader, follow closely the spelling of the manuscript
Introduction 5
Old Norse had nine basic vowel sounds, which might be long or short,
nasal or oral, giving 36 potential distinctions. Nasality seems to have
been lost in most people’s speech by about 1200, and so is ignored
here. Length is normally indicated by an acute accent. The relation-
ship between spelling and sound is as follows.
Notes:
There is no short counterpart of æ or long counterpart of ƒ. Both
sounds existed at one time, but in the kind of Old Norse on which the
normalised spelling is based short æ had coalesced with e and with
á. The use of œ to denote the long equivalent of ø is an arbitrary
convention, and in some works Õ is found.
Most of these sixteen distinctive vowels occur exclusively in stressed
syllables. In unstressed syllables there is no distinction of length and
for the most part a basic three-way contrast is found between a, i and
u. Some uncertainty exists about how these unstressed vowels were
pronounced, but the student will be safe enough using the following.
2.1.2 Diphthongs
may be falling (where the first element is stressed and the second un-
stressed, the latter usually a semi-vowel like English w in low or y in
say), rising (where the unstressed (semi)-vowel precedes the vowel as
in English with or yes) or balanced (where both elements are given
equal stress — as often in Faroese). Old Norse had three falling and a
great many rising diphthongs. However, because rising diphthongs
tend to be spelt with initial ‘j’ or ‘v’ in most forms of Scandinavian,
they are often regarded simply as sequences of j (as in English yes) or
v + vowel. This is more a theoretical than a practical question. Here
only the three falling diphthongs are listed separately. They were all
long and were pronounced as follows:
1. What is a diphthong?
2. What is the difference between a falling and a rising diphthong?
3. Which are the falling diphthongs of Old Norse?
4. Pronounce: lauss, bein, hey.
2.1.3 Consonants
Just as the vowels, so Old Norse consonants too may be long or short.
Consonants with prolonged articulation are not a normal feature of
English, but are heard in compounds, e.g. lake-country, pen-knife,
where the k and n sounds are extended. Consonant length in Old Norse
is indicated by gemination (doubling). The relationship between spell-
ing and sound is as follows.
Notes:
f Pronunciation (1) occurs in initial position (i.e. at the begin-
ning of words), pronunciation (2) in medial and final position
(i.e. in the middle or at the end of words).
g Pronunciation (1) occurs in initial position and immediately
12 Pronunciation and orthography
2.1.4 Syllables
Radical changes affected the sound system of Icelandic during the late
medieval period. This means that normalised Old Norse orthography
is not the best of guides to modern Icelandic pronunciation. Even so,
the correspondence between the two is far more regular than between
current spoken and written English.
Stress in modern Icelandic falls without exception on the initial syl-
lable of a word; in compounds the first syllable of the second element
has a strong secondary stress, e.g. spákona ‘prophetess’, with primary
stress on spá-, secondary on -kon-. Intonation can only sensibly be
learnt from listening to native speakers, or recordings of connected
speech, and will not be described here. The equivalents of Icelandic
sounds in other languages suggested below should be understood as
rough approximations.
2.2.1 Vowels
Notes:
a Pronunciation (2) occurs immediately before ng, (3) imme-
diately before gi, (1) in all other positions.
e Pronunciation (2) occurs immediately before ng, gi and gj,
(1) in all other positions.
i Pronunciation (2) occurs immediately before ng and gi, (1)
in all other positions.
o Pronunciation (2) occurs immediately before gi, (1) in all
other positions.
u Pronunciation (2) occurs immediately before ng, (3) imme-
diately before gi, (1) in all other positions.
ƒ Pronunciation (2) occurs immediately before ng and gi, (1)
in all other positions.
Although some of the above examples show long and others short
realisations of the different sounds, all vowels (except the diphthon-
gal variants of o and u, which are always short) may have either length.
Corresponding to long á in ár, for example, we have short á in árs
‘year [gen. sg.]’ (and also in langr, although written ‘a’); and corre-
sponding to short é in léttr we have long é in lét ‘let’, ‘caused’ (1st/3rd
sg. past of láta).
16 Pronunciation and orthography
2.2.2 Consonants
Notes:
f Pronunciation (1) occurs in initial position (i.e. at the begin-
ning of words), pronunciation (2) in medial and final position
(i.e. in the middle or at the end of words); (3) occurs immedi-
ately before l and n (except where a consonant follows, in which
case fl may be pronounced as in English flat before a voiceless
and as in English naval before a voiced consonant, and fn as m
(e.g. fíflt ‘seduced [pp.]’, fífldi ‘seduced [3rd sg. past]’, hefndi
‘avenged [3rd sg. past]’).
g Pronunciation (1) occurs initially before á, a, ó, o, ú, u, ø, ƒ, au
and consonants, medially before l and n (e.g. sigla ‘[to] sail’)
and also between consonants and a or u (e.g. saurga ‘[to] dirty’),
and finally after consonants (e.g. fling); pronunciation (2) occurs
initially before e, í, i, ‡, y, æ, œ, ei, ey and j, and medially between
consonants and i or j (e.g. helgi ‘holiness’); (3) occurs before s
and t (e.g. hugsa ‘[to] think’); (4) occurs between vowels and
a, u, r or ›, and finally after vowels; (5) occurs between vowels
and i or j.
gg Pronunciation (1) occurs between vowels and a, u, r or v, and
in final position; (2) occurs between vowels and i or j; (3)
occurs before t.
h Pronunciation (2) occurs before é and j, (3) before v, (1) in all
other positions.
k Pronunciation (2) occurs before e, í, i, ‡, y, æ, œ, ei, ey and j, (3)
before s and t, (1) in all other positions.
kk Pronunciation (2) occurs between a vowel and i or j, (3) before
s and t, (1) in all other positions; preaspiration means that a
Modern Icelandic 19
1. Of the written geminates bb, gg, ll, mm, nn, tt, which always de-
note long consonants in modern Icelandic?
2. How many different sounds can f denote, and what are they?
3. How many different sounds can g denote, and what are they?
4. Work through all the examples in 2.2.2, pronouncing each several
times.
2.2.3 Syllables
The fact that vowel length is regulated by the length of following con-
sonants means that in modern Icelandic there are effectively only two
types of stressed syllable, both long:
originally form part of the words in which it is now found, and be-
cause of its relatively late arrival, the epenthetic vowel is not indi-
cated in normalised Old Norse orthography. It develops between a
consonant (other than r) and r, especially an r in final position. Thus
ON ma›r ‘man [nom. m. sg.]’, dapr ‘sad [nom. m. sg.]’, eitr ‘poison
[nom./acc. n. sg.]’, bindr ‘tie(s) [2nd/3rd sg. pres.]’, for example, are
pronounced ma›ur, dapur, eitur, bindur (the first three with long
stressed vowels because only a single consonant immediately follows)
— and so written in modern Icelandic orthography.
Morphology deals with the form and structure of words, and syntax
with the ways in which words are combined to form sentences. In
section 3 we shall be concentrating on inflexional morphology (changes
in word-form that express grammatical categories and relationships,
sometimes called accidence) and the ways in which it interrelates with
syntax. In dealing with a language like Old Norse, where grammatical
categories such as number, gender, case, person, tense (see below
and 3.2, 3.6.1, 3.6.2) are expressed by variation in word-form, it is
unhelpful to divide the inflexions from the syntax, as has been com-
mon practice in earlier grammars. The student needs to appreciate from
the outset that form and function are interlinked: the form has no pur-
pose other than to express the function, and often the function cannot
be expressed without the form.
3.1.1 Number
3.1.2 Case
how they are, or may be, expressed, or the nature of the link between
the means of expression and the thing expressed.
In modern English a few words change form according to their func-
tion in a sentence. Thus we say (as a complete sentence):
I saw him
but
He saw me
not:
*Me saw he
or
*Him saw I
in which the roles are reversed, but it is the word-order that signals the
function (as it does additionally in I saw him and he saw me) not the
particular forms of the words involved, which do not change. Another
common means of expressing syntactic relationship in English is by
the use of function words (words which have little or no meaning on
their own) such as of, with, than. In:
24 Morphology and syntax
for example, of England modifies king, in much the same way as would
the addition of the adjective English. In:
than me supplies the part of the comparative phrase that denotes the
entity with which the comparison is made.
Where English uses word-order or function words to indicate syn-
tactic relationship, Old Norse regularly uses changes of word-form
instead or as well. This means that not only pronouns, but nouns and
adjectives (and also verbs and adverbs, as to some extent in English),
are likely to change form according to their relationship to other parts
of the sentence. It is their form that — wholly or partly — specifies
their grammatical role, as with I/me in English. English has traces of
such a system in the -’s (singular) or -s’ (plural) that may be added to
nouns. Instead of saying (or writing) the king of England as above, for
example, we may alternatively use England’s king; instead of the
comfort of passengers, passengers’ comfort. However, the Old Norse
system is vastly more complex than anything in English. Its heavy
reliance on form to indicate a variety of functions means that a simple
two-way distinction like that between English I and me or England
and England’s offers a wholly inadequate parallel.
Old Norse nouns, adjectives and pronouns exhibit four distinctive
case-forms, known as nominative, accusative, genitive and dative.
This means that a noun (or adjective or pronoun) potentially has eight
different actual forms (four in the singular, four in the plural), but in
reality most have fewer because the same form occurs in more than
one case.
The number of functions expressed by these case-forms greatly ex-
ceeds four. This means that no case is uniquely associated with a par-
ticular function: each is used in a variety of ways. The accusative, for
Noun inflexions and their function 25
3.1.3 Gender
As well as number and case, the role of gender in the inflexion of Old
Norse nouns needs to be considered. Gender is an inherent category
of the noun, that is, it is only when a noun is modified or referred to
that its gender becomes manifest. There is, for example, nothing about
the word dalr ‘valley’ to show that it is masculine rather than feminine
or neuter, but if it is modified by an adjective, that adjective will appear
in the appropriate masculine form, e.g. djúpr dalr ‘deep valley’ where
djúp- is the root of the word and -r the nom. m. sg. ending (see 3.3.4).
Similarly, if we wish to refer to a valley as ‘it’, it must be by the mas-
culine form of the personal pronoun: hann ‘he’.
While there is thus nothing gender-specific about any individual
Old Norse noun in its dictionary form, it is nevertheless true that gender
plays a part in the inflexional system of nouns, if only a minor one.
Most masculines, for example, end in -r or -i in the nominative sin-
gular, and many feminines in -a; neuters are characterised in both
singular and plural by a lack of distinction between nominative and
accusative, and many have no specific nom./acc. pl. inflexion either
(so that kvæ›i ‘poem’, for example, may be nom. or acc. sg. or pl.).
However, given that none of the above features (except the nom./acc.
sg./pl. identity of neuters) is totally restricted to one particular gender,
they cannot be classed as gender markers in the same way as the forms
of modifying adjectives or of anaphoric pronouns (pronouns that refer
back to some previously expressed meaning, as, for example, it refer-
ring to valley above). What the features do offer is guidance about the
likely gender of a noun — a useful insight since it can help (a) to see
which words in a sentence belong together and (b) to predict what
forms a given noun will have other than the particular one encountered.
28 Morphology and syntax
1. What does the grammatical category number refer to? Give exam-
ples from Old Norse.
2. What does the grammatical category case refer to? Give examples
from Old Norse.
3. What does the grammatical category gender refer to? Give exam-
ples from Old Norse.
4. To what extent does case occur in English?
5. In what ways other than change in word-form can syntactic rela-
tions be expressed?
6. Which cases are found in Old Norse, and how do we recognise
them?
nine (f.) and neuter (n.), this gives us six basic sets of endings. They
are as follows (~ = zero, i.e. there is no ending, the form consisting of
the root of the noun alone — e.g. dal ‘valley’, acc. m. sg.; ( ) = the
ending does not always occur; actual paradigms are given in 3.1.8).
Strong masculine
Weak masculine
Strong feminine
Weak feminine
Strong neuter
Weak neuter
Eiríkr (strong nom. m. sg.) is the subject, konungsma›r (strong nom. m. sg.)
the subject complement; for both subject and subject complement the nomi-
native is almost always the case used. Subject is an extremely hard concept to
get to grips with; it is sometimes loosely defined as ‘what the sentence is
about’; where the verb denotes an action, the subject is often the agent, or
‘doer’ of the action. However, such definitions relate chiefly to meaning. Syn-
tactically subjects may be defined both in English and Old Norse as the first
noun phrase of a sentence in unmarked word-order (where ‘noun phrase’ means
a noun or pronoun with or without accompanying modifiers — e.g. John, she,
32 Morphology and syntax
Bardagi (weak nom. m. sg.) is the subject, the first noun phrase in the sen-
tence.
Konur (weak nom. f. pl.), modified by tvær, is the subject; it is the first noun
phrase in the sentence and the women perform the action denoted by the verb
vƒk›u. Lei›inu (strong dat. n. sg. + def. art. — see 3.1.9) does not function
here as a noun phrase, but is part of the preposition phrase yfir lei›inu, in
which the noun is governed (i.e. has its case determined) by the preposition
yfir (see 3.7, 3.7.4).
Vápn (strong nom. n. pl.) is the subject, the first noun phrase in the sentence;
whether the weapons are seen as the agent, or ‘doer’, of the action, will depend
partly on the wider context, partly on the analysis; normally a human agent
wields weapons and the weapons are thus the instrument, but they can also be
portrayed as agent.
Mat and drykk (both strong acc. m. sg.) are objects of the verb tekr. Like
subject, object is a hard concept to define; traditionally a distinction is made
between ‘direct object’, the goal of an action, and ‘indirect object’ the benefi-
ciary, as in: I sent Peter (indirect object) a letter (direct object), but such defi-
nitions have to do with meaning rather than syntax. Syntactically objects may
Noun inflexions and their function 33
be defined both in English and Old Norse as the second and third noun phrases
of a sentence in unmarked word-order, with the accusative commonly mark-
ing the direct and the dative regularly marking the indirect object in Old Norse,
second position the indirect and third position the direct object (by and large)
in English. Mat and drykk are both direct objects: they are the goal of the
action, and whereas direct objects regularly appear unaccompanied by indi-
rect objects, the reverse is very uncommon (cf. the impossibility of English *I
gave him). The direct objects appear here in the accusative, the most common
case for this function.
Rei›a (weak acc. m. sg.), modified by allan, is the second noun phrase in the
sentence and the direct object of the verb báru. Skipinu (strong dat. n. sg. +
def. art.) does not function here as a noun phrase, but as part of the preposition
phrase af skipinu, and its case is determined by the preposition af (see 3.7.3).
Orrustur (weak acc. f. pl.), modified by margar, is the second noun phrase in
the sentence and the direct object of the verb átti. Englandi (strong dat. n. sg.)
is part of the preposition phrase í Englandi, and its case is determined by the
preposition í (see 3.7.4).
Daga (strong acc. m. pl.) is governed by the preposition of, skála (weak dat.
m. sg.), + its modifier miklum, by the preposition í.
This is an idiomatic phrase, of which Old Norse has its fair share. Augu (weak
acc. n. pl.) is governed by the preposition í, which requires the accusative here
because a sense of motion is involved (contrast examples 7 and 8).
34 Morphology and syntax
Brúsi (weak nom. m. sg.) is the subject; it is the first noun phrase in the sen-
tence and the man bearing the name performs the action denoted by the verb
dval›isk. Hrí› (strong acc. f. sg.), modified by litla, is an adverbial phrase
expressing duration of time (it answers the question: ‘How long?’).
Dráps (strong gen. n. sg.), the second noun phrase of the sentence, is the
direct object of the verb hefndi; hefna is one of the few verbs that take a direct
object in the genitive. fiorgríms (strong gen. m. sg.) is an objective genitive,
that is, it corresponds to English ‘of fiorgrímr’ and presents fiorgrímr as the
object or goal of an action (cf. ‘NN killed fiorgrímr’).
¯ndvegissúlna (weak gen. f. pl.), modified by sinna, is the direct object of the
verb leita. It comes in an infinitive clause (i.e., we have the infinitive at leita
‘to seek’, but only an implied subject). A full sentence might run: fleir leitu›u
ƒndvegissúlna ‘they sought the high-seat posts’, in which the direct object
would be the second noun phrase.
Gunnarr (strong nom. m. sg.), the first noun phrase in the sentence, is the
subject; he does the staying. Nótt (strong acc. f. sg.), modified by eina, is an
adverbial phrase expressing duration of time. Sigrí›ar (strong gen. f. sg.) and
frændkonu (weak gen. f. sg.) + sinnar are subjective (possessive) genitives,
that is, they correspond to English ‘-’s’ and present Sigrí›r, the kinswoman,
as the owner of the house where Gunnarr stayed (cf. ‘NN owns the house’);
note that ‘house’ is not expressed in the Old Norse sentence, paralleling English
usage as above or in, e.g., I am at Peter’s.
Noun inflexions and their function 35
Ingólfr (strong nom. m. sg.) is the subject, the first noun phrase in the sen-
tence and what it is about. Landnámsmanna (strong gen. m. pl.), modified by
allra, is a genitive of type, that is, it corresponds to English ‘of the settlers’
and presents landnámsmenn as a type of which Ingólfr is a representative.
Vatns (strong gen. n. sg.), modified by nƒkkurs, is governed by the preposition til.
This is an adage, consisting of two sentences. Augu (weak nom. n. pl.), the
first (and only) noun phrase in sentence 1, is the subject of the verb leyna; the
eyes fail to perform the action denoted by the verb. Kona (weak nom. f. sg.) is
the subject of the verb ann, the first noun phrase in sentence 2 and what it is
about. Manni (strong dat. m. sg.) is the direct object of ann, the second noun
phrase; a good many verbs take a direct object in the dative.
¯l (strong acc. n. sg.) is the direct object of the verb bera; it is the goal of the
action and the second noun phrase in the sentence. Víkingum (strong dat. m.
pl.) is the indirect object of bera; it denotes the beneficiary of the action and is
the third noun phrase. In English the indirect object may be expressed by a
preposition phrase (‘to the vikings’) or word-order (‘She was to serve the
vikings beer’ — indirect object before direct); in Old Norse the indirect object
appears in the dative.
Rƒgnvaldi (strong dat. m. sg.) is the indirect, fylg› (strong dat. f. sg.), modi-
fied by traustri, the direct object of hétu. As noted in connection with (16),
many verbs take a direct object in the dative, and heita ‘promise’ is among these.
Jarli (strong dat. m. sg.) is the indirect, lífs (strong gen. n. sg.) the direct
object of ljá. As noted in connection with (11), a few verbs take a direct object
in the genitive, and ljá ‘grant’ is among these.
Beinum (strong dat. n. pl.), modified by smám, has instrumental sense; in Old
Norse people are conceived as throwing with something (cf. the close seman-
tic relationship between English: He threw water onto the ground and He
splashed the ground with water). Instrumentality in Old Norse is expressed
either by the dative on its own or by the preposition me› ‘with’ + dat. Gólfit
(strong acc. n. sg. + def. art.), modified by flvert, is governed by the preposi-
tion um.
Dyrr (nom. f. pl.) is the first noun phrase in the sentence and the subject. It has
only plural forms, although it corresponds to the English singular ‘doorway’.
Noun inflexions and their function 37
These forms are also in part irregular (see 3.1.7.2, 3.1.7.4), and the nom. f. pl.
is indicated by other means than the adding of -a, -i or -u + r to the root. Enda
(weak dat. m. sg.) is governed by the preposition á.
we can tell from the agreement between the noun form hrí› and the
adjective form litla that we are dealing with the accusative feminine
singular. ‘Agreement’ means that there is a formal relationship be-
tween the two words, expressed by their having the same case, gender
and number (see further 3.3.1), and since litla can only be acc. f. sg. or
acc. m. pl. (see 3.3.9, paradigm 9), and hrí› cannot be acc. m. pl., the
case, gender and number they have in common must be acc. f. sg.
Having established that, we may further deduce that accusative case
in conjunction with a verb meaning ‘stayed’ (dval›isk) — and given
the sense of hrí› (‘while’ ‘short time’) — indicates duration of time.
In (6):
it is the form of the agreeing adjective, allan ‘all’, which shows that of
the various cases rei›a might be, singular or plural, it is in fact accusa-
tive singular, the adjective ending -an denoting acc. m. sg. alone (see
3.3.4). That it is accusative means it is likely to be the direct object of
the verb báru ‘carried’ (examination of the other words in the sen-
tence and their forms will in fact show that to be the only possible
analysis). That it is singular is of little consequence, since rei›i does
not normally appear in the plural (any more than ‘tackle’ in English).
That it is masculine is of importance to the extent that when used with
a function that requires accusative case, the agreeing adjective will
show the case, gender and number unambiguously.
Noun inflexions and their function 39
A further problem for the learner of Old Norse is that the endings
listed so far are by no means the whole story (cf., e.g., dyrr, nom. f.
pl., in example sentence 22). To be reasonably sure of recognising a
particular case-form for what it is, the student needs to be aware of
additional features that play their part in noun inflexion. These will
now be examined.
The rule that root a changes to ƒ before u should not cause the learner
problems, as long as s/he remembers that an unknown word with
ƒ in the root and u in the ending must be looked up in a dictionary
as though it had root a if it cannot be found there with root ƒ. For
example, stjƒrnur (nom./acc. pl.), fjƒru (acc./gen./dat. sg.), dƒlum,
grƒnnum (both dat. pl.) will be found not under stjƒ-, fjƒ-, dƒ-, grƒ-,
but under stjarna ‘star’, fjara ‘shore’, dalr ‘valley’, granni ‘neighbour’.
Somewhat greater difficulties are caused for the learner by the fact
that strong feminine and neuter nouns may exhibit root ƒ in certain
forms even though no u or v follows. The reason for the occurrence of
ƒ here is the presence of a following u at an earlier stage of the language.
This u caused a to develop to ƒ and was subsequently lost (e.g.
nom./acc. n. pl. *landu > lƒnd ‘countries’). The forms concerned are:
nominative, accusative and dative feminine singular and nominative
and accusative neuter plural. In addition, a small group of masculine
nouns (several of them very common) has root ƒ in the nominative
and accusative singular. If the feminine or masculine nouns are met
with in forms with root ƒ, no problem arises for the learner, since the
nominative singular is also the entry form in dictionaries. For neuter
plurals with root ƒ, however, root a must be substituted before
the word is sought in a dictionary, e.g. fjƒll ‘mountains’, lƒnd ‘coun-
tries’ will be found under fjall, land (see paradigm 26).
Difficulties with masculines and feminines of this type occur where
a form other than one with root ƒ is encountered. The bulk of the
feminines are the least troublesome: in the genitive singular, nomina-
tive, accusative and genitive plural these have root a, so the process of
looking such words up is simply the reverse of that which applies in
the case of those like stjƒrnur, fjƒru, dƒlum, or grƒnnum. For exam-
ple, kvalar (gen. sg. or nom./acc. pl.), hafnar (gen. sg.), hafnir (nom./
acc. pl.), kvala, hafna (gen. pl.) will be found not under *kval or *hafn,
but under kvƒl ‘torment’, and hƒfn ‘harbour’ (see paradigm 12). The
residual feminines and the masculines with original root a present a
more complex picture in that it is only in the genitive singular and
plural that a appears. In the remaining forms, where there is or has
been no u in the endings (automatically triggering a > ƒ, cf. above),
i.e., dative singular and nominative (also analogically accusative) plural
masculine, nominative and accusative plural feminine, we most often
find root e, though sometimes i (see 3.1.7.2). Thus vallar (gen. sg.),
Noun inflexions and their function 41
valla (gen. pl.), velli (dat. sg.), vellir (nom. pl.) should all be looked up
under vƒllr ‘field’, m., fjar›ar (gen. sg.), fjar›a (gen. pl.), fir›i (dat.
sg.), fir›ir (nom. pl.) under fjƒr›r ‘fjord’, m., strandar (gen. sg.),
stranda (gen. pl.), strendr (nom./acc. pl.) under strƒnd ‘beach’, f. (see
paradigms 4, 5, 18).
U-mutation affects unstressed as well as stressed syllables. In un-
stressed syllables, however, it results in u, thus enabling the mutation
to spread further. The strong masculine noun fƒgnu›r ‘joy’, for exam-
ple, comes ultimately from *fagna›uz via the intermediate forms, first
*fagnƒ›uz and then *fagnu›uz. The u-mutated forms — nom. and acc.
sg. and dat. pl.: fƒgnu›r, fƒgnu›, fƒgnu›um — thus contrast markedly
with the other parts of the paradigm whose root is fagna›- (see para-
digm 6). A strong neuter noun like sumar ‘summer’ has nom. and acc.
pl. sumur (< *sumuru < *sumƒru < *sumaru) (paradigm 27).
influence of earlier -ge, -ke (as in *dage, which became degi, see below)
— as well as analogical levelling (the restructuring of forms by the
force of analogy) — also played their part in this process. All the con-
ditioning sounds are likely to have been pronounced with the front of
the tongue raised close to its maximum height, and their presence had
the effect of turning a preceding back vowel (one pronounced with the
back of the tongue raised or lowered) into its front counterpart. Thus
ó, for example, which is a mid-high back vowel (the back of the tongue
is raised to above mid-height, but not to its full extent), became œ, a
mid-high front vowel, when an i, j or other ‘conditioner’ followed.
That is why the masculine noun fótr ‘foot’ and the feminine bók ‘book’
have nominative plurals fœtr, bœkr respectively (see paradigms 7 and
19).
As these introductory remarks and examples suggest, front muta-
tion, unlike the labial variety, is very much a historical process (it is
also common to most Germanic languages, cf. English foot — feet,
man — men, German Fuß — Füße, Mann — Männer). It occurred at
a stage of Scandinavian language development that preceded Old
Norse, and had ceased to be productive some time before the Old Norse
period. This has two important consequences for the recognition of
inflexions. First, we find an unstressed i that does not cause front mu-
tation because it arose after the period when mutation was taking place,
e.g. dat. m. sg. armi ‘arm’ (< *arme). This circumstance makes it
impossible to formulate a hard-and-fast rule (like a > ƒ before u, v)
stating which stressed vowels we can expect to find immediately pre-
ceding i. Second, the i, j or other conditioner triggering the fronting
may no longer be present (very often it is not — cf. fœtr and bœkr
above, earlier forms of which were *fotiz, *bokiz). This latter situa-
tion is parallel to the loss of u in forms such as fjƒll ‘mountains’, hƒfn
‘harbour’, noted in 3.1.7.1.
With such complications, what the learner of Old Norse needs to
know are the front mutation products of the back vowels affected, so
that s/he may recognise that fótr — fœtr or bók — bœkr, for example,
are different forms of the same lexical item. It is further useful to
know where in different paradigms to expect front-mutated root vowels.
The back : front correspondences arising from front mutation, to-
gether with examples (contrasting nom. sg. with nom. pl. unless other-
wise stated), are as follows:
Noun inflexions and their function 43
(7) dat. sg., gen. and dat. pl. of four nouns of relationship: fa›ir
‘father’, m.; bró›ir ‘brother’, m.; mó›ir ‘mother’, f.; dóttir ‘daughter’, f.
(e.g. fe›r, dat. sg., fe›ra, gen. pl., fe›rum, dat. pl.; mœ›r, dat. sg.,
mœ›ra, gen. pl., mœ›rum, dat. pl.; the dative singular of these nouns
can also be found with the ending -ur, causing labial rather than front
mutation in fa›ir; paradigms 8, 22).
It is further worth noting a small group of weak masculines consist-
ing of root + -and- suffix. These have front mutation in the nom. and
acc. pl., but it affects the suffix only (e.g. dómandi ‘judge’, nom. and
acc. pl. dómendr; with contraction of the suffix: bóndi (< búandi)
‘farmer’, nom./acc. pl. bœndr; paradigm 10).
3.1.7.3 Breaking
Certain endings occur that do not accord with those given on pp. 29–30.
We have already seen in 3.1.7.1 and 3.1.7.2 above that the nom. and
acc. pl. of a number of common masculine and feminine nouns end in
-r rather than -a, -i or -u + r. Other deviations which may cause prob-
lems of recognition are:
(1) Nominative masculine singulars that lack the -r ending. These
are due to the assimilation of r to an immediately preceding l, n, s (e.g.
lykill ‘key’, m. (< *lykilr); hrafn ‘raven’, m. (< *hrafnn < *hrafnr);
áss ‘god’, m. (< *ásr); see paradigms 3, 5). Hrafn exemplifies a general
rule that consonant + geminate (double) consonant is simplified to
consonant + single consonant (thus also in nouns with consonant +
root r; compare, e.g., nom. and acc. m. sg. vetr ‘winter’, the former
from earlier *vetrr). Nominative and accusative plural -r can be
assimilated in the same way as nom. sg. when the pl. ending does not
contain a vowel (e.g. mús ‘mouse’, f. — nom. and acc. pl. m‡ss).
(2) A small group of strong feminines that has -r in the nom. sg., just
as most strong masculines, and -i in the acc. and dat. sg. (e.g. hei›r,
hei›i, hei›i ‘moor’; see paradigm 17).
(3) Strong feminines with the suffix -ing or -ung, as well as a few
others, that have -u in the dat. sg. (e.g. dróttning ‘queen’, dat. sg.
dróttningu; sól ‘sun’, dat. sg. (usually) sólu; see paradigm 14).
(4) A small group of weak feminines that has -i throughout the
singular (e.g. gle›i ‘joy’; see paradigm 24). These nouns denote ab-
stract concepts and have no plural form.
(5) A few nouns with root nn that have -›r in the nom. sg. (e.g. mu›r
‘mouth’, m., acc. sg. munn; forku›r ‘strong desire’, f., gen. sg.
forkunnar).
The inflexions of ON nouns exhibit yet other deviations from the basic
pattern, but these are less likely to cause the learner problems of
recognition.
(1) The unstressed syllables of many disyllabic nouns lose their vowel
when an inflexional ending is added which itself consists of a syllable
(e.g. flistill ‘thistle’, m. — dat. sg. flistli, nom., acc., gen., dat. pl. flistlar,
46 Morphology and syntax
(15) ey ‘island’
Sg. nom. ey Pl. nom. eyjar
acc. ey acc. eyjar
gen. eyjar gen. eyja
dat. eyju dat. eyjum
(16) á ‘stream’
Sg. nom. á Pl. nom. ár
acc. á acc. ár
gen. ár gen. á
dat. á dat. ám
(20) tá ‘toe’
Sg. nom. tá Pl. nom. tær
acc. tá acc. tær
gen. tár gen. tá
dat. tá dat. tám
Noun inflexions and their function 51
Identify the case, gender, number, syntactic function and semantic role
of the nouns printed in bold in the following passage (adapted from
Hrafnkels saga Freysgo›a ‘The Saga of Hrafnkell, Priest of Freyr’).
Where a noun appears in a case other than the nominative singular,
give the nominative singular — the dictionary entry — form. In the
case of compound nouns, give the case, gender, number, function and
role of the last element only (e.g., in Brei›dal, analyse -dal, in
Hallfre›arstƒ›um, -stƒ›um and in fjárskiptis, -skiptis).
It was in days of-Haraldr king the hairfair, Hálfdan’s son the black, that
that man came with-ship REFL. POSS. to Iceland into Brei›dalr, who Hallfre›r
was-called.
It was in the days of King Haraldr fairhair, son of Hálfdan the black, that
a man called Hallfre›r brought his ship to Iceland, to Brei›dalr.
fiar var á skipi kona hans ok sonr, er Hrafnkell hét. Hann var fimmtán
vetra gamall. Hallfre›r setti bú saman.
There was on ship wife his and son, who Hrafnkell was-called. He was
fifteen of-winters old. Hallfre›r put dwelling together.
On board the ship was his wife and son, who was called Hrafnkell. He
was fifteen years old. Hallfre›r established a farmstead.
54 Morphology and syntax
But in spring-the moved Hallfre›r dwelling REFL. POSS. north over moor
and made dwelling there that is-called in Geitdalr.
But in the spring Hallfre›r moved his dwelling northwards across the moor
and made a dwelling in the place called Geitdalr.
And one night dreamt him that man came to him and said: ‘there lie you,
Hallfre›r, and rather unwarily. Move you a(-)way dwelling your and west
over Lagarfljót. There is fortune your all.’
And one night he dreamt that a man came to him and said: ‘There you lie,
Hallfre›r, and rather unwarily. Move your dwelling away and westwards
across Lagarfljót. There is where all your good fortune lies.’
Eptir flat vaknar hann ok fœrir bú sitt út yfir Rangá í Tungu, flar sem
sí›an heitir á Hallfre›arstƒ›um, ok bjó flar til elli.
After that wakes he and moves dwelling REFL. POSS. out over Rangá into
Tunga, there that later is-called at Hallfre›arsta›ir, and lived there till old-age.
After that he wakes up and moves his dwelling out across Rangá to Tunga,
to the place which has since been called Hallfre›arsta›ir, and lived there
into his old age.
En honum var› flar eptir geit ok hafr. Ok inn sama dag, sem Hallfre›r
var í brott, hljóp skri›a á húsin, ok t‡ndusk flar flessir gripir, ok flví
heitir flat sí›an í Geitdal.
But to-him came-to-be there behind she-goat and billy-goat. And the same
day that Hallfre›r was a(-)way, ran landslide onto houses-the, and lost-sk
[see 3.6.5.3] there these animals, and therefore is-called it since in Geitdalr.
But it turned out he left a she-goat and a billy-goat there. And the same day as
Hallfre›r moved away, a landslide fell onto the buildings and these animals
perished there, and for that reason the place has since been called Geitdalr.
Noun inflexions and their function 55
Hrafnkell lag›i flat í van›a sinn at rí›a yfir á hei›ar á sumarit. fiá var
Jƒkulsdalr albygg›r upp at brúm.
Hrafnkell laid that in custom REFL. POSS. to ride over onto moors in summer-
the. Then was Jƒkulsdalr fully-settled up to bridges.
Hrafnkell made it his practice to ride up onto the moors in the summer. At
this time Jƒkulsdalr was fully settled right up to the (rock) bridges.
And so when Hrafnkell came home, he asked his father for a division of
the property, and said he wanted to build a dwelling for himself.
fietta veitir fa›ir hans honum, ok hann gerir sér bœ í dal fleim, ok
kallar á A›albóli.
This grants father his to-him, and he makes for-self farm in valley that
and calls at A›alból.
His father grants him this, and he makes himself a farm in that valley and
calls it A›alból.
Masculine
Feminine
Neuter
The presence or absence of the initial -i (in some texts -e) is unlikely
to cause the learner serious problems of recognition. In the example
sentences in 3.1.5 we had on the one hand (3), (6) lei›inu, skipinu,
with dat. n. sg. -i noun ending + -nu, and on the other (20) gólfit with
acc. n. sg. zero noun ending + -it, all of them unambiguously combi-
nations of noun and definite article (cf. also vár-it, sumar-it (both acc.
n. sg.) and hús-in (acc. n. pl.) in the extract from Hrafnkels saga above).
Contrastive examples with and without -i, based on the list of noun
paradigms in 3.1.8, are:
Noun inflexions and their function 57
The most general rule governing the occurrence of initial -i in the def.
art. is that it is found in conjunction with words of one syllable and
omitted elsewhere (contrast the left and right-hand lists above). How-
ever, there are several exceptions to this.
The -i is omitted after the following monosyllabic forms.
(1) Nom./acc. m./f. pl. (cf. fœtr-nir (nom. m. pl.), fœtr-na (acc.
m. pl.) dœtr-nar (nom./acc. f. pl.)); an exception to the ex-
ception is represented by menn-inir, menn-ina (nom. and acc.
m. pl. respectively), though this is a rare type.
(2) Those ending in a vowel, but only where the article is disyl-
labic (contrast kné-in with á-nni above).
(3) Dative masculine singulars that lack the usual -i ending (cf.
sta›-num).
The -i is retained after genitive singulars in -ar (cf. sta›arins above,
further eyjarinnar (gen. f. sg.)).
Note that in the dative plural, the noun ending -(u)m loses its m and
the article is suffixed onto the u or stressed vowel (cf. kƒtlunum,
sƒgunum, or›unum, knjánum).
As with the nouns, certain regularities will be observed in the defi-
nite article paradigms. It will also be noticed that there are various
points of similarity between noun and article endings.
(1) The dat. pl. always ends in -um (as with nouns).
(2) The gen. pl. always ends in -a (as with nouns).
(3) It is only in the f. sg. and m. pl. there is a difference between
nom. and acc. forms.
(4) The gen. m. and n. sg. ends in -s, the gen. f. sg. in -ar (as with
most nouns, though some masculine genitives end in -ar).
(5) The nom. m. pl. ends in -ir, the acc. m. pl. in -a, and the
nom./acc. f. pl. in -ar (cf. the pattern for nouns: nom. m. pl.
and nom./acc. f. pl. = vowel + r, acc. m. pl. = vowel alone).
As well as the -(i)nn suffix dealt with here, Old Norse has a free-
58 Morphology and syntax
Identify the case, gender, number, syntactic function and semantic role
of the definite nouns printed in bold in the following sentences, and
insert a hyphen between noun and article. Where the noun appears in
a case other than nominative singular, give the nominative singular
definite form.
or persons s/he is addressing (Eng. you, sg. or pl.), 3rd person, in which
a person or persons other than the speaker/writer him/herself or the
one/those s/he is addressing are referred to (Eng. he, she, it, they).
nom. —
acc. sik
gen. sín
dat. sér
sá ‘that’, ‘those’
m. f. n.
Sg. nom. sá sú flat
acc. flann flá flat
gen. fless fleir(r)ar fless
dat. fleim fleir(r)i fl(v)í
The neuter singular and all the plural forms of this pronoun will be
seen to be identical with those of the personal pronoun, third person,
given in 3.2.1. In fact we are dealing with one and the same word. The
change in meaning from, for example, ‘those female beings’ or ‘those
feminine objects’ to ‘they [f.]’ is very small. Indeed, the same develop-
ment can be observed in many languages (French il ‘he’, elle ‘she’, for
example, come from the Latin pronoun ille ‘that’), and on occasion
Old Norse sá, sú are found in place of hann, hon.
Although there is considerable irregularity in the paradigm, com-
pare the acc., gen., dat. m. sg. endings -nn, -ss, -m and the acc., gen.,
dat. f. sg. -á, -ar, -i with those of the corresponding forms of the suf-
fixed definite article (3.1.9) and of hinn immediately below.
64 Morphology and syntax
m. f. n.
Sg. nom. hinn hin hitt
acc. hinn hina hitt
gen. hins hinnar hins
dat. hinum hinni hinu
The student will observe the close similarity between the forms of
this pronoun and those of the suffixed definite article (though note the
-tt in the nom./acc. n. sg.). There is in fact a strong likelihood that the
suffixed article is a reduced form of hinn. Not only does the similarity
of form suggest this, the development: demonstrative pronoun > defi-
nite article is quite widely attested (cf., for example, French le, la —
like il, elle, though by a different route — from Latin ille ‘that’). On
the relationship between hinn and (h)inn, the free-standing definite
article of Old Norse, see 3.3.5.
m. f. n.
Sg. nom. sjá/flessi sjá/flessi fletta
acc. flenna flessa fletta
gen. flessa flessar/flessar(r)ar flessa
dat. flessum/fleima flessi/flessar(r)i flessu/flvísa
singulars sjá and flessi are both common, but dat. m. sg. fleima and
dat. n. sg. flvísa are much less so. The genitive and dative feminine
singular and the genitive plural can be thought of as flessar, flessi,
flessa respectively, basic forms which are sometimes expanded by suf-
fixes (flessar-(r)ar, flessa-r(r)a), or infixes (fless-ar(r)-i). Although
the forms of this pronoun may appear anomalous, similarities with
other paradigms can still be found. The endings of the plural in par-
ticular are very close to those of hinn (above), and even in the singular
we notice the characteristic -a and -ar endings in the feminine accusa-
tive and genitive, and -um, -i, -u in the masculine, feminine and neuter
dative respectively. Some of the remaining forms also show charac-
teristic features, though not in the endings — observe the n, t and s of
acc. m. -nn-, nom./acc. n. -tt- and gen. m. and n. -ss-.
m. f. n.
Sg. nom. nƒkkurr nƒkkur nƒkkut
acc. nƒkkurn nƒkkura nƒkkut
gen. nƒkkurs nƒkkurrar nƒkkurs
dat. nƒkkurum nƒkkurri nƒkkuru
In the oldest sources many of the forms of this pronoun appear with
root nakkvar-.
Other indefinite pronouns are einnhverr ‘some(one/thing)’, and sumr
‘some’. The former consists of an invariable ein-, except in the
nom./acc. m. and n. sg. (einn-, eitt- respectively), and occasionally the
gen. m. and n. sg. (eins-), + hverr, the inflexions of which are de-
scribed below. The latter inflects like a strong adjective (see 3.3.4).
Of the sundry negative pronouns of Old Norse the only one the learner
will encounter regularly is engi ‘no one’ ‘nothing’ ‘none’ ‘no’. The
various forms of the other negatives, manngi ‘no one’, vættki ‘nothing’,
hvárigr or hvárgi ‘neither’, will, when met with, be well enough under-
stood from the glosses and examples given in Old Norse dictionaries.
m. f. n.
Sg. nom. engi engi ekki
acc. engan/engi enga ekki
gen. enskis engrar enskis
dat. engum engri engu
The paradigm presented here gives the most common forms of engi.
A complete list of attested forms will be found in Noreen 1923 (p. 323).
Virtually all of these are easily deducible, however, as long as it is
known (1) that the root of the word may be eing- or øng- as well as
eng-, and (2) that -v- may be added before endings beginning with -a
(e.g. nom./acc. f. pl. øngvar, engvar) and before the -ir of the nom. m.
pl. (e.g. øngvir, engvir).
The inflexions of engi present a familiar enough pattern (observe,
however, nom./acc. n. sg. ekki, from *eitt-ki < *eitt-gi). The student
Pronoun inflexions and their function 67
should compare the endings given above with those of hinn and
nƒkkurr, especially the latter, and make a note of where they coincide.
Only forms peculiar to engi need be learnt specially.
The two principal pronouns in this category are hverr ‘who’ ‘what’
‘which’, ‘each’ ‘every’, and hvárr ‘which of two’, ‘each of two’ (sg.),
‘which of two groups’, ‘each of two groups’ (pl.). With the exception
of the acc. m. sg. forms, hvern and hvárn, both decline like strong
adjectives (see 3.3.4). In common with some adjectives hverr inserts
a -j- between root and endings beginning with -a or -u; hvárr does not.
For ease of overview, the complete paradigm of hverr is now given.
m. f. n.
Sg. nom. hverr hver hvert
acc. hvern hverja hvert
gen. hvers hverrar hvers
dat. hverjum hverri hverju
Ek (1st person sg. nom.) is the subject; ‘I’ is the agent or ‘performer’ of the
action and the first noun phrase in the sentence. fiér (2nd sg. dat.) is the indirect
object; ‘you’ is the beneficiary of the action and the second noun phrase. fiat
(3rd. sg. n. acc.) is the direct object, the goal of the action (i.e. what is said)
and the third noun phrase.
The subject is fleir (3rd pl. m. nom.), the agent and first noun phrase in the
sentence. fieim (3rd pl. dat.) does not function here as a noun phrase, but is
part of the preposition phrase at fleim, in which the pronoun is governed (i.e.
has its case determined) by the preposition at (see 3.7, 3.7.3).
The subject is -tu (2nd sg. nom., suffixed to the verb); it is the agent and first
noun phrase in the sentence. Nƒkkut is part of the direct object. The second
noun phrase and the goal of the action consists of the noun li›sinni (acc. n.
sg.) modified by the pronoun nƒkkut (which since it appears here in the role of
modifier functions adjectivally; see 3.2). As a modifier nƒkkut appears in the
same case (acc.), gender (n.) and number (sg.) as its head word (li›sinni). This
formal relationship between the two (whereby the head word determines the
form of its modifier) is known as grammatical agreement or concord and is a
regular phenomenon in Old Norse (see 3.3.1). Okkr (1st dual dat.) is the indirect
object; it denotes the beneficiary of the action and is the third noun phrase in
the sentence.
This example consists of two sentences. Hann (3rd sg. m. acc.) in sentence 1
is part of the preposition phrase yfir hann, and its case is determined by the
preposition yfir (see 3.7, 3.7.4). Hann (3rd sg. m. nom.) in sentence 2 is sub-
ject, the first noun phrase and the agent. fieira (3rd pl. gen.) is part of the
preposition phrase me›al fleira, and its case is determined by the preposition
me›al (see 3.7.2).
70 Morphology and syntax
(7) fiví skaltu heita mér, at koma aptr til mín at ƒ›ru hausti
‘That shall-you promise to-me, to come back to me at second
autumn’
‘You must promise me to come back to me next autumn’
fiví (3rd sg. n. dat.) is the anticipatory direct object: the thing promised is ‘to
come back . . . ’, but the infinitive clause — the equivalent of a noun phrase —
is postponed and its place filled by the pronoun flví. The unmarked position
for the direct object would be somewhere after the subject and the finite verb
(skaltu), but here it has been moved to the front of the sentence for emphasis.
The subject is -tu (2nd sg. nom., suffixed to the verb); it is the agent, and —
the fronted flví apart — the first noun phrase in the sentence. Mér (1st sg. dat.)
is the indirect object, the beneficiary (the person to whom the promise is made),
and the second or third noun phrase. Mín (1st sg. gen.) is part of the preposi-
tion phrase til mín, and its case is governed by the preposition til (see 3.7.2).
Pronoun inflexions and their function 71
ver›r (sá er ver›r X ‘that person is worthy of X’) to the front of the sentence
(cf. the identical movement in English); its case is determined by the adjec-
tive ver›r (cf. English worthy/deserving of something), and its neuter gender
by the fact that it does not refer to anything of specifically masculine or femi-
nine gender. In traditional analysis y›r (2nd pl. dat., but used here as a singular
honorific, like French vous — the person being addressed is the king) would
be classed as the indirect object (the recipient or experiencer of the ‘seeming’,
cf. ‘to-you’), but recently claims have been made for the existence of a class
of ‘oblique’ (i.e. non-nominative) subjects into which y›r here would fall
(note that with flyk(k)ir ‘seems’ the person to whom something seems is
normally always the first noun phrase in the sentence; see further 3.9.3). Sá
(nom. m. sg.) is the subject of the elliptical infinitive clause (sá [vera] ver›r
‘that man [to be] worthy’) — what the clause is about. fietta rá› (acc. n. sg.),
with fletta modifying rá›, is the direct object of the dependent sentence, the
goal of the action (the words refer to the thing given or proffered); it is the
only noun phrase in the sentence, the subject being subsumed into the relative
particle or complementiser er, which is best regarded as being outside the
sentence (see 3.8, 3.8.2.1).
This example consists of two sentences. In sentence 1, hin vistin (nom. f. sg.),
with hin modifying vist-in (noun + def. art.), is the subject; it is the ‘performer’
of the action and the first noun phrase. In sentence 2, sjá (nom. f. sg.) is also
the subject, fulfilling on its own the same function as hin vistin in sentence 1.
Notice how hin contrasts with sjá: ‘that other one’ as opposed to ‘this one’.
Hon (3rd sg. f. nom.) is the subject, the agent and the first noun phrase in the
sentence. fienna (acc. m. sg., referring to an entity — person, animal or object
— of masculine gender) is the direct object (what is valued) and the second
noun phrase. Hinn (acc. m. sg., likewise referring to an entity of masculine
gender) is part of a comparative phrase; this can be understood as ‘more than
[she valued] the other’, and hinn taken as a direct object too.
Hann (3rd sg. m. nom.) is the subject, not the agent here but the experiencer,
and the first noun phrase in the sentence. The noun phrase flessara tí›inda
(gen. pl.), with flessara modifying tí›inda, has its case determined by the
adjective varr (cf. English aware of something).
This example consists of two sentences, in both of which sumir (nom. m. pl.)
is subject, the first noun phrase and, in sentence 2, the agent. In sentence 1
with its passive verb phrase (váru drepnir ‘were killed’, see 3.6.4) the subject
is the recipient or goal of the action (a typical feature of passive construc-
tions).
Hon (3rd sg. f. nom.) is subject, the agent and first noun phrase in the sen-
tence. Engu (dat. n. sg.) can be construed as the direct object of svarar (what
is answered), but in origin it probably had instrumental sense (the idea of
answering with something, cf. 3.1.5, sentence 20).
Engi skip (nom. n. pl.), with engi modifying skip, is subject, the (potential)
‘performer’ of the action and the only noun phrase in the sentence.
phrase) and fronted (cf. (11) above); the sense is instrumental, hence the use
of the dative. It will be observed that (19) is without a subject (i.e. there is no
element that corresponds to X in the abstraction above); although rare in
English, subjectless sentences are a regular feature of Old Norse (see 3.9.3).
(22) Nú forvitnar mik at vita, hverja ek hefi hér fóstrat, e›r hverrar
ættar flit eru›
‘Now interests me to know whom I have here fostered, or of-
what family you-two are’
‘Now I am curious to know whom I have been fostering here,
or what family you two belong to’
This example consists of three sentences and an infinitive clause (at vita).
Sentence 1 is what is traditionally called ‘impersonal’, by which is meant that
it has no nominative subject; such an analysis would class mik (1st sg. acc.) as
direct object. More recent approaches would see mik as an oblique subject (cf.
Pronoun inflexions and their function 75
(11) and (21) above), the experiencer and first (and only) noun phrase. Hverja
(acc. m. pl.) in sentence 2 is the direct object — the goal of the action — but
since it takes the form of an interrogative pronoun, it is fronted from its un-
marked position after subject and finite or non-finite verb (cf. ek hefi fóstrat
hann or ek hefi hann fóstrat). Ek (1st sg. nom.) is the subject of sentence 2, the
agent and, apart from the fronted interrogative, the first noun phrase. In hverrar
ættar (gen. f. sg.) in sentence 3, with hverrar modifying ættar and the whole
phrase fronted because of the presence of the interrogative, the genitive has a
defining or connective sense (note that once again the Old Norse genitive can
correspond to English of). fiit (2nd dual nom.) is the subject: the X of an X is
Y construction, and, the fronted interrogative apart, the first noun phrase.
Hann (3rd sg. m. nom.) is subject: the X of an X is Y construction and the first
noun phrase. Hverjum manni (dat. m. sg.), with hverjum modifying manni, is
the second proposition in a comparative construction — the proposition de-
noting the entity with which the comparison is made (cf. 3.1.5, sentence 21);
the dative phrase is the equivalent of the noun phrase X (in whatever case is
appropriate) that follows en ‘than’ in a comparative adjective + en construc-
tion (e.g. fleiri en X ‘more than X’).
Hvern (acc. m. sg.) does not function here as a noun phrase, but is part of the
preposition phrase fyrir hvern, and its case is determined by the preposition
fyrir (see 3.7.4). Hinna (gen. pl.) has partitive sense: ‘the others’ is the whole
of which each individual is a part (cf. (20) above).
76 Morphology and syntax
3.3.2 Definiteness
son flinn (acc. m. sg.) ‘the youngest son your [i.e. your youngest son]’.
Sometimes where used as an epithet a weak adjective may occur with-
out a determiner, e.g. Eiríkr rau›i ‘Eiríkr the red’; here the adjective
alone carries the definite sense ‘the red’.
Adjectives in Old Norse, together with adverbs, are inflected for degree.
There are three degrees: positive, comparative and superlative, corre-
sponding in form to English: big — bigger — biggest. As in English,
the positive degree has no special inflexion, and therefore the form of
an adjective in the positive is simply its root plus the appropriate
inflexion to indicate number, case, gender and definiteness. The com-
parative and superlative degrees are normally marked by the suffixes
-(a)r, -(a)st respectively; to the superlative suffix is added the appro-
priate strong or weak ending just as in the positive, to the comparative
suffix a limited range of endings that indicate number, case and gender
(see 3.3.4 below). Comparative and superlative forms of the adjective
are thus double-inflected, e.g. hvass-ar-i (comp. nom. m. sg., f. sg.,
nom./acc./gen. pl.) ‘sharper’, d‡r-r-a (comp. acc./gen./dat. m. sg.,
n. sg.) ‘dearer’ ‘more precious’, hvass-ast-ar (sup. strong nom./acc.
f. pl.) ‘sharpest’, d‡r-st-a (sup. strong acc. f. sg., acc. m. pl., weak
acc./gen./dat. m. sg., nom. f. sg., n. sg.) ‘dearest’ ‘most precious’.
Just as in the case of noun inflexion (see 3.1.4), it is the basic patterns
the student needs to grasp. Minor variations — to the extent they cause
problems of understanding — can be noted and learnt when they are
encountered.
From 3.3.2 and 3.3.3 above it will be clear that — the comparative
and superlative suffixes and comparative endings apart — there are
two distinct types of adjective inflexion in Old Norse, strong and weak.
Both types, as already observed, inflect for number, case and gender.
With two numbers, four cases and three genders, there is thus a possi-
ble total of twice twenty-four different inflexions. In fact, because the
same form may occur in different parts of the paradigm, the total is
much smaller: fundamentally, there are thirteen different strong ad-
jective forms and just four weak. They are as follows (~ = zero, i.e.
there is no ending, the form consisting of root alone — e.g. rík ‘pow-
erful’, strong nom. f. sg.; actual paradigms are given in 3.3.9).
Strong masculine
Weak masculine
Strong feminine
Weak feminine
Strong neuter
Weak neuter
The student should further observe the close similarity between the
strong adjectival endings and those of pronouns such as hinn, nƒkkurr,
engi, hverr (cf. 3.2.2, 3.2.3, 3.2.4, 3.2.5). The similarity becomes even
clearer when the many adjectives with an -in suffix are added to the
equation and the comparison is extended to certain of the possessive
adjectives and the suffixed definite article (probably a reduced form
of hinn, cf. 3.2.2).
Adjectives in -in inflect according to the tables above, but with three
distinct deviations (see the example kominn, 3.3.9, paradigm 7).
(1) Where the tables show an ending in or beginning with -r, adjectives
in -in have -n instead, e.g. -inn (strong nom. m. sg.), -inni (strong dat.
f. sg.). This is because an earlier r has assimilated to the n (-inn < *-inr,
-inni < *-inri, cf. hinn < *hinr, 3.2.3). (2) The n of the suffix disap-
pears in the strong nom./acc. n. sg. ending, giving -it (the end result of
the development *-int > *-itt > -it, cf. hitt, nom./acc. n. sg. of hinn
(3.2.2)). (3) The strong acc. m. sg. has the same form as the nom.,
ending in -inn. It should also be noted that the -i- of the -in suffix is
dropped when the inflexional ending consists of an additional sylla-
ble, except in the strong gen. and dat. f. sg. and the strong gen. pl., e.g.
-nir (strong nom. m. pl.), -ni (weak nom. m. sg.), -inna (strong gen.
pl.). This is not unlike what happens to the suffixed definite article
(see 3.1.9), although the pattern is not wholly identical. Most two-
syllable adjectives, in fact, drop the unstressed vowel of the second
syllable according to the pattern of those in -in. A great many of these
have an -al, -il, or -ul suffix (see the example gamall, 3.3.9, paradigm 8),
and, just as with the n of -in, the immediately following r of the in-
flexional endings is assimilated to the l, giving -ll(-) instead of the
expected *-lr(-), e.g. -all (strong nom. m. sg.), -allar (strong gen. f. sg.).
Adjective inflexions and their function 83
The possessive adjectives of the first and second person and the
third person reflexive possessive (i.e., words corresponding to English
‘my’, ‘our’, etc. and, with pronominal function, ‘mine’, ‘ours’, etc.)
inflect according to one or other of the strong adjective patterns just
discussed. Minn ‘my’ (see 3.3.9, paradigm 21), flinn ‘your [sg.]’, sinn
‘his/her/its/their own’ go for the most part like adjectives in -in (but
without loss of the i at any point since in the possessives it is part of
the root syllable). It is worth noting, however, that in having the nom./
acc. n. sg. forms mitt, flitt, sitt, they parallel even more closely the
paradigm of the pronoun hinn, the only difference between the two
being that the root vowel of the possessives is long before everything
except a geminate consonant, e.g. minn (nom. m. sg.), míns (gen. m.
or n. sg.). Várr ‘our [pl.]’ is inflected according to the strong pattern
of the tables above, except that, as with certain pronouns, the acc. m.
sg. ends in -n (várn). Okkarr ‘our [dual]’, ykkarr ‘your [dual]’ and
y›(v)arr ‘your [pl.]’ parallel várr (acc. m. sg. okkarn, ykkarn, y›(v)arn),
but as two-syllable words drop the unstressed vowel of the second
syllable according to the pattern of the two-syllable adjectives dis-
cussed above (giving, for example, acc. f. sg. okkra, ykkra, y›ra).
It remains to list the adjective endings that follow the comparative
suffix.
Masculine
Feminine
Neuter
This minimal set of endings is also the one used with present parti-
ciples, e.g. sofandi (nom. m. sg., f. sg., nom./acc./gen. pl.) ‘sleeping’
(see 3.3.9, paradigm 19).
how the definite article manifests itself in noun phrases that include
adjectives.
As will have been apparent from certain of the examples in 3.3.2,
Old Norse has a free-standing definite article in addition to the suf-
fixed variety (just as in the modern Scandinavian languages). The free-
standing article occurs where a definite noun is modified by an adjective
(the adjective normally always being weak), e.g. (h)inn blindi ma›r
‘the blind man’. It is also used where an adjective with definite func-
tion (once again weak) is ‘substantivised’, i.e. used without a noun
and thus, in a sense, in its place, e.g. (h)inir au›gu ‘the rich [pl.]’. (The
inflexional forms of (h)inn are the same as those of the demonstrative
pronoun hinn given in 3.2.2, except for the nom./acc. n. sg. which is
(h)it with a single t. Note that in some texts instead of (h)inn, (h)it,
(h)inir, etc. we get enn, et, enir, i.e., no initial h- and root vowel e.)
In Old Norse prose neither of the constructions just illustrated is in
fact particularly common except where something or someone is being
distinguished from another or others of the same type or name, e.g.:
hin sí›asta orrosta ‘the last battle’, hinna gƒmlu skálda ‘the old poets
[gen. pl.]’ (as opposed to the new ones), hendi inni hœgri ‘the right
hand [dat. f. sg.]’, Óláfr inn helgi ‘Óláfr the saint’, hit sí›ara ‘the
latter’, hinn flri›i ‘the third’. (Observe that the free-standing article
and its accompanying weak adjective may be found either before or
after the noun).
To express the equivalent of English the + adj. ± noun Old Norse
employs a variety of other constructions. In prose a much more com-
mon rendering of the definite article than (h)inn on its own is (h)inn
together with the demonstrative pronoun sá (see 3.2.2), giving phrases
of the type: sá (h)inn blindi ma›r ‘that the blind man’, ma›r sá (h)inn
blindi ‘man that the blind’, or, less commonly, sá ma›r (h)inn blindi
‘that man the blind’, i.e. (in all three cases) ‘the blind man’. (Note the
possible variations in word-order, and that sá and (h)inn agree with,
i.e. always appear in the same case, gender and number as, adjective
and noun — here nom. m. sg.; see 3.3.1.) Occasionally (h)inn may be
omitted, and we then get the phrase-type: sá blindi ma›r or ma›r sá
blindi, where sá alone renders ‘the’. In Norwegian sources in particu-
lar, the suffixed article may be used in addition to its free-standing
counterpart, or the demonstrative sá, or both together, giving phrases
like hinn hvíti bjƒrninn ‘the white bear-the’ (literally), hƒndin sú hœgri
86 Morphology and syntax
‘hand-the that right’, sá hinn flƒgli ma›rinn ‘that the silent man-the’,
vápnin flau in gó›u ‘weapons-the those the good’, all equivalents of
English the + adj. + noun. (Note that the phrase-types without hinn,
e.g. sá blindi ma›r, hƒndin sú hœgri, sometimes have greater deictic
emphasis, i.e. the pronoun is closer in meaning to ‘that’ than ‘the’.)
Observe the fundamental identity of (h)inn and the suffixed definite
article, the former of which certainly and the latter probably derive
from demonstrative hinn (see above and 3.2.2). Additional notes on
word-order in noun phrases will be found in 3.9.2.
Identify the case, gender, number, syntactic function and semantic role
of the definite noun phrases printed in bold in the following sentences.
(5) Engi ma›r mátti nefna hann annan veg en jarl hinn illa
‘No man might call him another way than earl the bad’
‘No man might call him anything other than “the bad earl”’
(6) En flau hin stóru skip, er á›r hƒf›u siglt, ok fleir hug›u
at Ormrinn væri, flat var hit fyrra Tranan, en hit sí›ara
Ormr hinn skammi
‘But those the big ships which before had sailed, and they
thought that Ormrinn were, that was the former Tranan but
the latter Ormr hinn skammi’
Adjective inflexions and their function 87
‘But as for the big ships which had sailed previously and which
they thought were “The Serpent”, the former was “The Crane”
and the latter “The Short Serpent”’
As was done for nouns and pronouns, examples are now given of
adjectives in function. With the wide range of adjectival functions and
inflexions that exists, only a selection can be illustrated, with the
emphasis on the most common types. As far as is practicable, the ex-
amples are ordered as follows: (a) strong adjectives; (b) weak adjec-
tives; (c) substantivised adjectives (strong and weak); (d) superlatives
(strong and weak); (e) comparatives — though some sentences con-
tain examples of more than one type. In other respects, the exemplifi-
cation follows the same pattern as for nouns (see the preamble on
p. 31). Note that the adjectival inflexions being illustrated (or the whole
word where there is no difference from the root form) are printed in
bold type. To underline the grammatical relations involved, bold is
also used for the noun or pronoun with which the adjective agrees.
Compare the inflexions used below with those set out and discussed
in 3.3.4. Observe, too, the differences between Old Norse and English
phraseology and sentence formation. Definitions of basic concepts that
have already been given are not repeated; if in doubt the student should
consult the individual commentaries that accompany each of the ex-
amples of noun function in 3.1.5.
88 Morphology and syntax
Skakkr and kalla›r are nom. m. sg., agreeing with hann, the subject. Skakkr is
used predicatively (see 3.3), as the subject complement (i.e. as Y in: X is/
becomes/is called Y); it has indefinite function and therefore strong inflexion.
Kalla›r is the past participle of the verb kalla ‘[to] call’ which together with
var forms a passive phrase (see 3.6.4); in such phrases the past participle
(which itself functions not unlike a subject complement) inflects as a strong
adjective.
Mikil and hƒr› are nom. f. sg., agreeing with orrosta (f.), the subject. They are
attributive adjectives (see 3.3), occurring in an indefinite noun phrase and
therefore having strong inflexion. Although hƒr› as a nom. f. sg. strong adjec-
tive is without ending, the root vowel has u-mutation, just as the nom. sg. of
strong feminine nouns (see 3.1.7.1 and 3.3.8.1).
Kauplaust is nom. n. sg., agreeing with flat, the subject. It is the subject com-
plement; it has indefinite function and therefore strong inflexion. Eigi mun
flat kauplaust is elliptical for Eigi mun flat kauplaust vera (see 3.9.5.2).
Búnir is nom. m. pl., agreeing with fleir, the subject. It is the subject comple-
ment; it has indefinite function and therefore strong inflexion.
Skamma is acc. f. sg., agreeing with hrí› (f.), which is accusative because it
functions as a time adverbial (see 3.1.2 and 3.1.5, sentence 10). The adjective
is used attributively, and, occurring in an indefinite noun phrase, has strong
inflexion.
Vá›væna is acc. m. pl., agreeing with strauma (m.), which is the direct object.
The adjective is used attributively, and, occurring in an indefinite noun phrase,
has strong inflexion.
Handtekinn is acc. m. sg., agreeing with jarl (m.), which is the direct object.
The adjective is used predicatively, as the object complement; it has indefinite
function and therefore strong inflexion.
Sín is acc. n. pl., agreeing with segl (n.), which is the direct object of the
infinitive clause. Note that though the reflexive possessive agrees with flá
(the subject of vinda, cf. 3.9.4) in person (both are 3rd), it agrees with segl in
case, gender and number. Possessives have only strong forms: they are
themselves determiners, not part of what is determined or defined.
Stórra is gen. pl., agreeing with hƒggva (n.), the noun of the preposition phrase
til hƒggva stórra. The case of the noun is governed by the preposition til (see
3.7, 3.7.2). The adjective is used attributively, and, occurring in an indefinite
noun phrase, has strong inflexion.
90 Morphology and syntax
Y›varra is gen. pl., agreeing with verka (n.), which is an objective genitive,
that is, it corresponds to English ‘of . . . ’ and presents the deeds as being
rewarded (cf. ‘X rewarded the deed’). Note that though the possessive adjective
(here functioning as a reflexive, cf. 3.2.1) agrees with flér in person (they are
both 2nd pl.), it agrees with verka in case and number (gender is not marked
in the gen. pl.). (On the strong inflexion of y›varra, see (8) above.)
Allgó›um is dat. m. sg., agreeing with fagna›i (m.), the noun of the preposition
phrase í allgó›um fagna›i. The case of the noun is governed by the preposition
í. The adjective is used attributively, and, occurring in an indefinite noun phrase,
has strong inflexion.
Kærr is nom. m. sg., agreeing with A›ils (m.), the subject. It is the head word
of the subject complement; it has indefinite function and therefore strong
inflexion. Gó›um is dat. pl., agreeing with hestum (m.), the noun of the
preposition phrase at gó›um hestum. The case of the noun is governed by the
preposition at. The adjective is used attributively and has indefinite function
and therefore strong inflexion. The preposition phrase modifies the adjective kærr.
Adjective inflexions and their function 91
Drengiligi is nom. m. sg., agreeing with ma›r (m.), the subject (cf. who is X?
X is Y, where X is the subject). The adjective is used attributively, and, occur-
ring in a definite noun phrase, has weak inflexion.
Unga is acc. m. sg., agreeing with Eindri›a (m.), the direct object. It is used
as a ‘defining’ epithet (Eindri›i ‘the young’ as opposed to any other Eindri›i);
as such it is part of a definite noun phrase and therefore has weak inflexion.
Observe that definite function in itself is enough to trigger weak inflexion,
there being no determiners in the noun phrase in question.
(16) Hann var sonr Óláfs ins hvíta ok Au›ar innar djúpú›gu.
‘He was the son of Óláfr the white and Au›r the deep-minded’
Hvíta is gen. m. sg., agreeing with Óláfs (m.), and djúpú›gu gen. f. sg., agree-
ing with Au›ar (f.); both nouns are subjective (possessive) genitives (Óláfr
and Au›r have ‘him’ as their son). As in (15), the adjectives are used as
‘defining’ epithets, but here in conjunction with the free-standing article (h)inn.
Both are part of definite noun phrases and therefore have weak inflexion.
Sinni and flrœnzku are dat. f. sg., agreeing with ambótt (f.), which is the indi-
rect object of bau› (cf. that he gave an order to the bondwoman). Note that
though the reflexive possessive agrees with hann in person (both are 3rd), it
agrees with ambótt in case, gender and number. firœnzku is used attributively,
and, occurring in a definite noun phrase, has weak inflexion (on the strong
inflexion of sinni, see (8) above).
92 Morphology and syntax
Gauzku is acc. m. pl., agreeing with menn (m.), the direct object. It is used
attributively, and, occurring in a definite noun phrase, has weak inflexion.
Fáir is nom. m. pl. (masculine is the default gender where the reference is to
people in general). The adjective stands on its own without a noun and forms
the subject. It has indefinite function and therefore strong inflexion.
Jafna is acc. f. sg., agreeing with refsing (f.), which is the direct object of
hafa. The adjective is used attributively, and, occurring in an indefinite noun
phrase, has strong inflexion. Ríkan and óríkan are acc. m. sg. They stand on
their own without a noun — but referring to individual males — and form the
direct object of lét. They have indefinite function and therefore strong inflexion.
The construction here is what is known as an accusative and infinitive: the
accusative objects of lét, ‘mighty and unmighty’, are in a sense also the subjects
of hafa (see 3.9.4).
Illu is dat. n. sg. It stands on its own without a noun and is part of the prepo-
sition phrase frá illu, its case being determined by the preposition. Gott is acc.
n. sg., and it, too, stands on its own without a noun. It is the direct object of the
verb ger. Both adjectives have indefinite function and therefore strong
inflexion. They are neuter because they do not refer to an entity of masculine
or feminine gender. (Note that snústu is a contracted form of snúsk flú: -sk
verb (see 3.6.5.3) + the personal pronoun flú (literally ‘turn you’). Gott is an
irregular nom./acc. n. sg. form (see 3.3.8.4), nom. m. sg. gó›r.)
Adjective inflexions and their function 93
N‡komna is dat. m. sg. It stands on its own without a noun — but referring to
a male animal — and forms the head of the indirect object phrase flessum
enum n‡komna. Determined by flessum enum, it has definite function and
therefore weak inflexion.
Skygnastr is nom. m. sg. sup., agreeing with Sveinn (m.), the subject. The
adjective is the subject complement; it has indefinite function and therefore
strong inflexion (which follows the superlative -ast suffix). (Note that the
superlative here is what is known as absolute, i.e. it denotes not the highest
but a very high degree.) Allra is gen. pl. of allr ‘all’, which has only strong
forms; it agrees with manna, a genitive of type: menn are presented as a type
of which Sveinn is a particularly sharp-sighted member (see 3.1.5, sentence 14).
Frægsta is nom. f. sg. sup., agreeing with fer› (f.), the subject. It is the subject
complement; it has definite function and therefore weak inflexion (which fol-
lows the superlative -st suffix). (Note that here too the superlative is absolute
(see (23)) — notwithstanding the definiteness of the noun phrase.)
(25) Me›an hann var á léttasta aldri, haf›i hann hvert sumar
lei›angr úti
‘While he was at lightest age, had he each summer levy out’
‘While he was at the most active age, he made naval expedi-
tions each summer’
Léttasta is dat. m. sg. sup., agreeing with aldri (m.), the noun of the preposi-
tion phrase á léttasta aldri. The case of the noun is governed by the preposi-
tion á. The adjective is used attributively and has definite function and therefore
weak inflexion (which follows the superlative -ast suffix). On the occurrence
of weak inflexion in the absence of determiners, cf. (15).
94 Morphology and syntax
Bezti is nom. m. sg. sup. It stands on its own without a noun and forms the
subject complement. It has definite function and therefore weak inflexion.
Bezt- and the comparative betr- are suppletive forms (i.e. they have a different
root from other parts of the word, cf. positive gó›-; see further 3.3.8.3); ‘z’
denotes the sounds ts (cf. 2.1.3), so what we have is in effect *bet-st-. The
phrase hinn bezti is elliptical for hinn bezti kostr ‘the best choice/option’.
Fúsari is nom. m. sg. comp., agreeing with jarl (m.), the subject of the second
sentence. The adjective is the subject complement. Following the compara-
tive suffix -ar, we get the appropriate comparative inflexion, which remains
the same whether the function is indefinite or definite.
(28) fiar gekk Rƒgnvaldr jarl af skipum ok allt it gƒfgara li› fleira
‘There went Rƒgnvaldr earl off ships and all the more-noble
force their’
‘There Earl Rƒgnvaldr and all the more noble of their force
left the ships’
Allt is nom. n. sg. of allr ‘all’, which has only strong forms. Together with
gƒfgara (nom. n. sg.) it agrees with li› (n.), one of the two subjects. Both
adjectives are used attributively. In gƒfgara, following the comparative suffix
-ar, we get the appropriate comparative inflexion (see (27)). The comparative
and superlative forms gƒfgar-, gƒfgast- show loss of an unstressed vowel: the
positive root is gƒfug- (see 3.3.8.5 point (1)).
(29) Hin yngri skáld hafa ort eptir dœmum hinna gƒmlu skálda
‘The younger poets have composed following the examples
of the old poets’
Yngri is nom. n. pl. comp., agreeing with skáld (n.), the subject. The adjective
is used attributively. Following the comparative suffix -r, we get the appropri-
ate comparative inflexion (see (27)). The comparative and superlative forms
yngr- yngst- have a different root vowel from the positive ung- (because of
Adjective inflexions and their function 95
front mutation, see 3.3.8.2). Gƒmlu is gen. pl., agreeing with skálda (n.), a
possessive genitive (the examples, in a sense, ‘belong to’ the old poets). The
adjective is used attributively, and, occurring in a definite noun phrase, has
weak inflexion. The form gƒmlu has suffered loss of the second, unstressed,
syllable, and its root vowel has undergone u-mutation (strong nom. m. sg.
gamall; cf. 3.3.4, 3.3.8.5 point (1), 3.1.7.1, 3.3.8.1).
(30) fieir lƒg›u á flat hit mesta kapp, hverr betri hesta átti
‘They laid on that the most contest, who better horses owned’
‘They made it a matter of the greatest rivalry who owned the
better horses’
Mesta is acc. n. sg. sup., agreeing with kapp (n.), the direct object of lƒg›u.
The adjective is used attributively; it has definite function and therefore weak
inflexion (which follows the superlative -st suffix). The comparative and super-
lative forms meir-, mest- are suppletive (positive mikil-; see (26)). Betri is
acc. m. pl. comp., agreeing with hesta (m.), the direct object of átti. The ad-
jective is used attributively. Following the comparative suffix -r, we get the
appropriate comparative inflexion (see (27)). Like the superlative bezt-, betri
is a suppletive form (cf. (26)).
As in the case of nouns (cf. 3.1.6), the learner may initially experience
some difficulty in recognising which adjective inflexions are which.
The strong endings are by and large distinctive, and even where an
ending recurs in different parts of the paradigm there are unlikely to
be serious problems of understanding. Although the genitive mascu-
line and neuter singular, for example, both end in -s, they clearly sig-
nal the genitive singular, just as -ra is an unambiguous sign of the
genitive plural. The identity between nominative and accusative in
the neuter singular and feminine and neuter plural may be problematic,
but very often their function and therefore their case will be apparent
from the context.
It is when confronted with the minimal distinctions of the weak and
comparative systems of endings — and their overlap with certain strong
endings — that the learner will regularly have to rely on the presence
or absence of other words in the noun phrase, and, where appropriate,
their forms, to determine the case, gender and number of the adjec-
tive. Fortunately, as we have seen, it is the way of adjectives, and
weak adjectives in particular, to be accompanied by words with which
they exhibit grammatical agreement. In sentence (17), for example:
it can be shown that flrœnzku is dat. f. sg., even though one strong and
eleven other weak forms have the -u ending, because of the presence
of sinni and fleirri. These words determine the noun that flrœnzku modi-
fies, which means (a): the function of the adjective is definite and the
form weak, and (b): flrœnzku will have the same case, gender and
number as sinni and fleirri since all three agree with the noun ambótt.
Given that sinni and fleirri are unambiguously dat. f. sg., we can thus
deduce that flrœnzku represents the same combination of case, gender
and number. Similarly, in sentence (22):
it can be shown that n‡komna is dat. m. sg., even though two strong
Adjective inflexions and their function 97
and seven other weak forms have the -a ending. Here the noun phrase
lacks a noun with which the adjective can agree, but there is agree-
ment with the determiners flessum enum. Their presence indicates the
phrase is definite and the adjective therefore weak, and although flessum
enum can represent the dat. m. sg. or dat. pl., in combination with
n‡komna the pair can only be dat. m. sg. since the dat. pl. adjective
ending (weak and strong) is -um. In sentence (27):
unga might represent the strong acc. f. sg. or acc. m. pl., or the weak
acc., gen. or dat. m. sg. or n. sg. (any case). The only word with which
unga can agree, Eindri›a, is masculine singular. That excludes the
possibility of feminine or neuter gender, or masculine plural, but given
that weak nouns have exactly the same forms in the singular as weak
adjectives, it does not help determine whether unga is acc., gen. or
dat. (m. sg.). Here one has to rely on function. The verb drepa takes a
direct object in the accusative, and since its -r ending shows Erlingr to
be nominative and thus subject, and there are no other noun phrases in
the sentence, Eindri›a unga must be the direct object and therefore
accusative.
The basics of front mutation were discussed in 3.1.7.2. The only parts
of adjectival inflexion affected by this process are certain comparative
and superlative forms. Those adjectives that form the comparative with
the -r and the superlative with the -st suffix undergo front mutation of
back root vowels. We find the following back : front correspondences
(the examples contrast the strong nom. m. sg. pos. form with the nom.
m. sg. comp.):
100 Morphology and syntax
(b) The strong nom./acc. n. sg. of gó›r is gótt (cf. point (2) above)
or, much more commonly, gott, of margr mart (also sometimes margt)
and of sannr ‘true’ satt.
(c) The two-syllable adjective heilagr ‘holy’, which drops the a of
the unstressed syllable on the pattern of those in -al etc. (see 3.3.4 and
3.3.8.5 point (1)), regularly undergoes monophthongisation (i.e. the
diphthong ei changes to a single vowel) in the shortened forms, cf.,
for example, strong nom./acc. f. pl. helgar, weak nom. m. sg. helgi.
sg., f. sg., etc. comp. hær(r)i (with front mutation); see paradigms 5, 20).
(4) As with nouns (cf. 3.1.7.5 point (2)), the vowels of inflexional
endings tend to be dropped when they immediately follow a long vowel
of the same or similar quality (e.g. grár — strong acc. m. sg. grán,
strong dat. m. sg., dat. pl. grám, strong acc. f. sg., dat. n. sg. and acc.
m. pl., all weak forms except nom. m. sg. and dat. pl. grá; trúr ‘faith-
ful’ — strong dat. n. sg., weak acc., gen., dat. f. sg. and nom., acc.,
gen. pl. trú; see paradigms 5, 15, 20).
(5) As with nouns (cf. 3.1.7.5 point (4)), j is inserted in some adjectives
before inflexional endings or suffixes consisting of or beginning in a
or u; in others v is inserted before endings or suffixes consisting of or
beginning in a or i (e.g. n‡r ‘new’ — strong acc. m. sg. n‡jan, strong
dat. m. sg., dat. pl. n‡jum, strong acc. f. sg. and acc. m. pl., weak acc.,
gen., dat. m. sg., nom. f. sg. and n. sg. n‡ja, strong nom. m. sg. sup.
n‡jastr; døkkr ‘dark’ — strong acc. m. sg. døkkvan, weak nom. m. sg.
døkkvi, nom. m. sg., f. sg., etc. comp. døkkvari; see paradigms 6, 20).
(6) As a final irregularity, it is worth noting that a few adjectives
including the third person possessives are uninflected. These will cause
the learner little trouble, since, like adjectives in English, their form
remains unchanged whatever their function. Examples are: einskipa
‘with one ship’, fulltí›a ‘full-grown’, andvaka ‘sleepless’. Some of
these, e.g. andvaka, can also appear in the nom. m. sg. with an -i end-
ing: andvaki. The possessives are: hans ‘his’, hennar ‘her’, fless ‘its’,
fleir(r)a ‘their’, i.e. the genitive forms of the corresponding pronouns
which have taken on an additional, adjectival function.
Strong inflexion
Weak inflexion
Degree inflexion
(20) Main types, illustrated by strong and comp. nom. m. sg. forms
pos. comp. sup.
sjúkr ‘ill’ sjúkari sjúkastr
hreinn ‘pure’ hreinni hreinstr
hvass ‘sharp’ hvassari hvassastr
grár ‘grey’ grár(r)i grástr
fƒlr ‘pale’ fƒlvari fƒlvastr
hei›inn ‘heathen’ hei›nari hei›nastr
au›igr ‘wealthy’ au›igri au›gastr
Adjective inflexions and their function 111
(5) Hann spur›i, hverr fyrir skipi flví ré›i enu vegliga
‘He asked who over ship that commanded the magnificent’
‘He asked who commanded the magnificent ship’
(10) fieir fundu flegar sveininn flar sofanda hjá húsi einu
‘They found immediately boy-the there sleeping by house one’
‘They at once found the boy sleeping there beside a certain
building’
Adjective inflexions and their function 113
(12) Hann hélt ƒll heit sín drengiliga vi› sína menn
‘He kept all promises REFL. POSS. nobly with REFL. POSS. men’
‘He nobly kept all his promises to his men’
(15) Erlendr vildi ekki, at synir hans hef›i lægra hlut flar í
Eyjum
‘Erlendr wanted not that sons his had lower lot there in Islands’
‘Erlendr did not want his sons to have a poorer position there
in the Orkneys’
(21) Hann lét flar gera steinkastala gó›an; var flat ƒruggt
vígi
‘He let there make stone-castle good; was that secure fortress’
‘He had a fine stone castle made there; it was a secure fortress’
3.4 Numerals
Cardinal Ordinal
1 einn fyrstr
2 tveir annarr
3 flrír flri›i
4 fjórir fjór›i
5 fim(m) fim(m)ti
6 sex sétti
7 sjau sjaundi
8 átta áttandi, átti
9 níu níundi
10 tíu tíundi
11 ellifu ellifti
12 tólf tólfti
13 flrettán flrettándi
14 fjórtán, fjƒgurtán fjórtándi, fjƒgurtándi
15 fim(m)tán fim(m)tándi
16 sextán sextándi
17 sjaut(j)án sjaut(j)ándi
18 át(t)ján át(t)jándi
116 Morphology and syntax
19 nítján nítjándi
20 tuttugu tuttugundi
21 tuttugu ok einn tuttugundi ok fyrsti
22 tuttugu ok tveir tuttugundi ok annarr
30 flrír tigir flrítugundi
31 flrír tigir ok einn flrítugundi ok fyrsti
40 fjórir tigir fertugundi
50 fim(m) tigir fim(m)tugundi
60 sex tigir sextugundi
70 sjau tigir sjautugundi
80 átta tigir áttatugundi
90 níu tigir nítugundi
Beyond nítugundi the ordinals are not recorded. The cardinals are:
As is apparent from this list, hundra› may denote either 100 (hundra›
tírœtt ‘a hundred of ten tens’ (see below)) or 120 (hundra› tólfrœtt ‘a
hundred of twelve tens’ ‘a “long” hundred’). Correspondingly, flúsund
= either 1000 or 1200. With figures of this magnitude precision is
often unimportant in Old Icelandic literature; where the intention is to
express exact numbers or amounts, either the text itself or the edition
may make clear what is meant by hundra› or flúsund.
Many of the above numerals have variant forms (the most common
are specified), but none is likely to cause the learner problems of recog-
nition. It is worth noting that the ordinal suffix -undi also appears as
-andi (though níundi and tíundi are almost universal).
Tigir is nom. pl. of tigr (also tegr, tugr, tøgr), a masculine noun
meaning ‘a group of ten’ ‘a decade’; flrír tigir, for example, thus liter-
ally means ‘three tens’. Hundra› is a neuter noun, and flúsund a femi-
nine. All three inflect according to noun paradigms (tigr according to
paradigm 5 but without the vowel changes associated with root a,
hundra› according to 27 but without the loss of unstressed a in dat.
sg. and gen./dat. pl., flúsund according to 13).
Numerals 117
Of the other cardinals, only einn, tveir, flrír and fjórir inflect, einn
for case, gender, number and definiteness (cf., e.g., flat eina ‘the one’),
tveir, flrír and fjórir for case and gender alone. The paradigms are as
follows (only the strong forms of einn are given).
einn ‘one’
m. f. n.
Sg. nom. einn ein eitt
acc. einn eina eitt
gen. eins einnar eins
dat. einum einni einu
tveir ‘two’
m. f. n.
Pl. nom. tveir tvær tvau
acc. tvá tvær tvau
gen. tveggja tveggja tveggja
dat. tveim(r) tveim(r) tveim(r)
flrír ‘three’
m. f. n.
Pl. nom. flrír flrjár flrjú
acc. flrjá flrjár flrjú
gen. flriggja flriggja flriggja
dat. flrim(r) flrim(r) flrim(r)
fjórir ‘four’
m. f. n.
Pl. nom. fjórir fjórar fjƒgur
acc. fjóra fjórar fjƒgur
gen. fjƒgurra fjƒgurra fjƒgurra
dat. fjórum fjórum fjórum
118 Morphology and syntax
In the plural, einn has the sense ‘only’ ‘just’ (e.g. sagnir einar ‘just
tales’), ‘some’ (einir hei›nir víkingar ‘some heathen vikings’) and
‘roughly’ ‘about’ (einar flrjár ‘about three’).
As a numeral (in the sense ‘the two’) we can include bá›ir ‘both’. It
has only plural (dual) forms and inflects as follows.
m. f. n.
nom. bá›ir bá›ar bæ›i
acc. bá›a bá›ar bæ›i
gen. beggja beggja beggja
dat. bá›um bá›um bá›um
It will be seen that einn (in its strong form) inflects exactly like the
pronoun hinn (3.2.2); it is also very similar in its forms to the
possessives minn, flinn, sinn and adjectives with the -in suffix (3.3.9,
paradigms 21 and 7). Certain of the inflexions of tveir, flrír, fjórir and
bá›ir are reminiscent of corresponding noun, pronoun and adjective
forms. Compare (1) nom. m. tveir with nom. m. pl. fleir ‘they’ ‘those’,
and flrír, fjórir, bá›ir with the common nom. m. pl. ending -ir;
(2) acc. m. tvá, flrjá with acc. m. pl. flá ‘them’ ‘those’, and fjóra, bá›a
with the common acc. m. pl. ending -a; (3) nom./acc. f. tvær with
nom./acc. f. pl. flær, and flrjár, fjórar, bá›ar with the common nom./
acc. f. pl. ending -ar; (4) gen. tveggja, flriggja, fjƒgurra, beggja with the
ubiquitous gen. pl. ending -a; (5) dat. tveim, flrim, fjórum, bá›um with
the ubiquitous dat. pl. ending -(u)m. Observe also that the nom./acc. n.
of all four numerals, like the nom./acc. n. pl. of nouns, pronouns and
adjectives, is without a final added -r.
The ordinals all inflect as adjectives (3.3.4), but subject to consider-
able restrictions: fyrstr has both strong and weak inflexion (contrast
hann gekk fyrstr ‘he went first’ and hit fyrsta sumar ‘the first sum-
mer’), annarr only strong, and the remaining ordinals only weak; flri›i
has j insertion before -a and -u (3.3.8.5 point (5)). The paradigm of
annarr shows a number of irregularities and is therefore given here
in full.
Numerals 119
m. f. n.
Sg. nom. annarr ƒnnur annat
acc. annan a›ra annat
gen. annars annarrar annars
dat. ƒ›rum annarri ƒ›ru
1. What types of numeral are found in the basic Old Norse counting
system?
2. Which of the numerals are nouns?
3. In what way may hundra› and flúsund be ambiguous?
4. Which of the cardinal numbers inflect, and what points of similar-
ity are there between their paradigms and those of other word classes
in Old Norse?
5. Which of the ordinal numbers inflect, and according to which
pattern or patterns?
120 Morphology and syntax
Inflexions, where they occur, are printed in bold (or the whole word,
where the inflected form is identical with the root). Notes explain the
relationship between form and function, and usage in general. Compare
the inflexions used with those set out or identified above. Observe,
too, the differences between Old Norse and English phraseology and
sentence formation. Definitions of basic concepts that have already
been given are not repeated; if in doubt, the student should in the first
instance consult the commentaries that accompany each of the exam-
ples of noun usage (3.1.5).
(1) fiat var tíunda dag jóla, er Rƒgnvaldr jarl stó› upp
‘It was tenth day of-Christmas that Rƒgnvaldr earl got up’
‘It was on the tenth day of Christmas that Earl Rƒgnvaldr got up’
Tíunda is an ordinal number in the acc. m. sg., agreeing with dag, which is
accusative because the phrase of which it is the head is a time adverbial (it
answers the question: ‘When?’; see 3.1.2). Note that jól ‘Christmas’ is a plural
noun.
(2) Sigur›r jarl átti flrjá sonu a›ra; hét einn Sumarli›i, annarr
Brúsi, flri›i Einarr rangmunnr
‘Sigur›r earl had three sons other; was-called one Sumarli›i,
second Brúsi, third Einarr rangmunnr’
‘Earl Sigur›r had three other sons; one was called Sumarli›i,
the second Brúsi, the third Einarr wry-mouth’
firjá is a cardinal and a›ra an ordinal number; both are acc. m. (pl.), agreeing
with sonu, the direct object of the first sentence. Einn is a cardinal and annarr
and flri›i are ordinal numbers; all three are nom. m. sg., the subjects of their
respective sentences (cf. ‘ . . . the second [was called] . . . the third [was
called] . . .’). Annarr, unlike almost all other ordinals, has strong inflexion.
(3) Af herfangi flví, er vér fám flar, skulum vér fá fátœkum mƒnnum
inn fimmtuganda penning
‘Of booty that which we get there, shall we give to-poor men
the fiftieth penny’
‘Of the booty which we win there, we will give the fiftieth
part to the poor’
Numerals 121
Fimmtuganda is an ordinal in the acc. m. sg., agreeing with penning, the direct
object.
Bá›um is a plural numerical adjective in the dat., agreeing with endum, the
noun of the preposition phrase á bá›um endum. The case of the noun is gov-
erned by the preposition á (see 3.7, 3.7.4).
Fjóra is a cardinal in the acc. m., agreeing with the numerical noun tigu, the
direct object. Together, fjóra and tigu make up the numeral ‘forty’. Numbers
which include or consist of the nouns tigr, hundra› or flúsund are followed by
a genitive of type, that is, one which expresses the nature of the entity to
which the numeral refers (cf. English hundreds of people).
Tólfrœtt is an adjective in the strong nom. n. sg., agreeing with the numerical
noun hundra›, the subject. On the genitive manna, see (5) above.
(7) fiá er hann var fimtán vetra gamall, rei› hann til flings
‘Then when he was fifteen of-winters old, rode he to assembly’
‘When he was fifteen years old, he rode to the assembly’
Tólfta is an ordinal in the acc. m. sg., agreeing with mann, the noun of the
preposition phrase vi› tólfta mann. The case of the noun is governed by the
preposition vi› (see 3.7.4). ‘With ORDINAL man’ is a common way in Old Norse
of specifying the total number in a group: the subject (here: ‘he’) is included
in the group and the ordinal gives the total. In this particular example there
were twelve altogether: ‘he’ and eleven others.
Tveir is a cardinal in the nom. m., agreeing with menn, the subject. Fimta is an
ordinal in the gen. m. sg., agreeing with tigar; ins fimta tigar is a partitive
genitive, that is, one that expresses the whole of which the number or num-
bers singled out for mention (here: ‘two men’) are a part (see 3.2.6, sentence
20). Note how the number is expressed: the fourth ‘ten’ ends at 40 and the
fifth at 50, thus two of the fifth ten, i.e. two on the way from 40 towards 50, is 42.
(11) Rƒgnvaldr jarl gaf Haraldi jarli annat skipit; flat hét Fífa, en
annat hét Hjálp
‘Rƒgnvaldr earl gave to-Haraldr earl the-one ship-the; it
was-called Fífa, and the-other was-called Hjálp’
‘Earl Rƒgnvaldr gave Earl Haraldr the one ship; it was called
Arrow and the other was called Help’
Annat is an ordinal. In the first sentence it appears in the acc. n. sg., agreeing
with skipit, the direct object. In the third sentence it is nom. n. sg. and is the
subject. Note that annarr not only means ‘second’, ‘other’, but ‘the one’ and
‘the other’ of two entities. Observe further that skipit is definite (i.e. it is ac-
companied by the (suffixed) definite article); this gives the construction parti-
tive sense (cf. (9) above), i.e. it means literally ‘the one of the [two] ships’.
Numerals 123
3.5 Adverbs
Adverbs are sometimes defined as words that modify the verb or specify
its action (e.g. English quickly in he ran quickly, which denotes the
manner of the running). In fact, the function of many words that are
traditionally classed as adverbs does not fit this definition. Indeed the
adverb word class serves as a kind of dustbin into which items that do
not obviously belong to any other category can be put. In the English
sentence: unfortunately, they could not come, for example, unfortu-
nately is classed as an adverb, yet it says nothing about the ‘coming’,
but means rather: ‘I/we think it is unfortunate they could not come’.
As in English, adverbs in Old Norse are a heterogeneous group.
One feature they all share, however, is that they do not inflect for
number, person, case or gender. Most, like adjectives, inflect for degree
(see below), but that is all. The learner therefore needs to master nothing
more than the meanings of adverbs and to be able to recognise their
comparative and superlative forms.
parts formed with an -i suffix that denote rest in a place and others
with an -an suffix meaning movement from a place (e.g. inn ‘in’, inni
‘inside’, innan ‘from within’). Some have only two of the forms (e.g.
nor›r ‘northwards’ ‘in the north’, nor›an ‘from the north’ (with root
nor›-); thus also the other compass-point adverbs — on su›r ‘south-
wards’, sunnan ‘from the south’, see p. 126). Certain of these locational
adverbs can have special meanings (e.g. útan ‘from without’ and thus
‘from Iceland’, seen from the perspective of Norway). The threefold
distinction: ‘to a place’, ‘in a place’ and ‘from a place’ is also found in
the commonly occurring hingat — hér — he›an ‘hither — here —
hence’, flangat — flar — fla›an ‘thither — there — thence’, hvert —
hvar — hva›an ‘where (to) — where — whence’ (it should be noted
that none of these words has the slightly archaic ring of some of the
English equivalents). Adverbs with the -an suffix combine with a pre-
ceding fyrir (cf. 3.7.1, 3.7.4) to form prepositional phrases indicating
position relative to another (fixed) position (e.g. fyrir nor›an hei›ina
‘north of the heath’, fyrir ofan húsin ‘above the buildings’; note the
idiomatic fyrir nor›an/sunnan land ‘in the north/south of Iceland’).
Other minor deviations from the above pattern include the addition
of an extra r to many comparatives in -ar (e.g. optar(r) ‘more often’,
framar(r) ‘farther forward’ ‘farther on’ from fram ‘forward’) and the
spread of comparative r into many superlatives (e.g. frama(r)st ‘far-
thest forward’ ‘farthest on’, inna(r)st ‘farthest in’). As the brackets in
the examples indicate, regular forms may also be found.
Like adjectives, adverbs that form the comparative with the -r and
the superlative with the -st suffix undergo front mutation of back root
vowels (see 3.1.7.2 and 3.3.8.2). Few adverbs are in fact affected.
Common ones are lengra ‘farther’, lengst ‘farthest’ (see above), fremr
‘farther forward’, fremst ‘farthest forward’ (alternative comparative
and superlative forms to framar(r), frama(r)st, see above), and the
irregular fjarri ‘far off’ — firr ‘farther off’ — first ‘farthest off’ and
gƒrva ‘thoroughly’, ‘precisely’ — gørr ‘more thoroughly’ — gørst
‘most thoroughly’.
Also in common with adjectives, a small group of adverbs have
suppletive forms in the comparative and superlative (see 3.3.8.3). The
ones likely to be encountered regularly by the learner are:
gjarna —— heldr —— helzt ‘willingly’, ‘rather’, ‘most of all’
illa —— verr —— verst ‘badly’, ‘worse’, ‘worst’
lítt ——minnr —— minst ‘little’, ‘less’, ‘least’
mjƒk —— meir(r) —— mest ‘much’, ‘more’, ‘most’
vel —— betr —— bezt ‘well’, ‘better’, ‘best’
One or two of the deviations and minor irregularities affecting ad-
jectives (3.3.8.4, 3.3.8.5) can be found in the comparative and/or su-
perlative forms of adverbs also. Attention has already been drawn to
the loss of neuter -t in superlatives of the skjótast type (3.3.8.5 point
(2)). Further to be noted are consonantal assimilations (3.3.8.4 point
(1), cf., e.g., seinna ‘more slowly’ ‘later’ < *seinra) and loss of un-
stressed syllables in disyllabic adverbs when a further syllable is added
(3.3.8.5 point (1), cf., e.g., sjaldan ‘seldom’ — sjaldnar ‘more sel-
dom’ —sjaldnast ‘most seldom’).
Occasionally the alternation ›r — nn seen in nouns like mu›r
‘mouth’, acc. munn, gen. munns etc. (cf. 3.1.7.4 point (5)) also distin-
guishes different forms of adverbs. We have already noted that the -an
form of su›r is sunnan; its comparative and superlative manifesta-
tions are sunnar(r), sunna(r)st respectively.
Adverbs 127
(7) Hann eignar sér svá allar eyjar fyrir vestan Skotland
‘He assigns to-self thus all islands west of Scotland’
‘He thus takes possession of all the islands west of Scotland’
Svá is an adverb of manner; it refers to the way in which ‘he’ takes possession
of the islands. Note the compound preposition fyrir vestan incorporating the
adverb of place vestan (see 3.5.1).
For flá, see (3) above. S‡sligast is an adverb of manner in the superlative; it
describes the way in which fiór›r worked and defines it as the highest degree
of that manner of working. Although allra is in form the gen. pl. of the adjec-
tive allr, it functions here as an intensifying adverb: fiór›r did not just work
most briskly, but most briskly of all (things).
In the above we have dealt almost exclusively with adverbs, that is,
single words whose basic function is adverbial and which are there-
fore assigned to the adverb word class. However, as litlu and allra in
example sentences (11) and (12) above indicate, non-adverbs can some-
times be used in such a way that they assume adverbial function. This
applies not only to single words, but to whole phrases. Many preposi-
tion phrases, for example, are reducible to adverbs. Thus í + a place-
name is in a sense the equivalent of flar ‘there’ or hér ‘here’ in that it
answers the question ‘where’ (e.g. í Nóregi ‘in Norway’), and í + a noun
denoting a point in time or a period of time is the equivalent of flá,
answering the question ‘when’ (e.g. í fleim tíma ‘at that time’). Noun
phrases, too, may have adverbial function (e.g. flat var einn dag ‘that
happened one day’ where einn dag is accusative and expresses a point
in time; cf. 3.1.2 and 3.1.5, sentence 10). Even complete dependent
sentences may be reducible to a single adverb and thus be shown to
have adverbial function (e.g., me›an hann lif›i ‘while he lived’ refers
to a period of time and can be replaced by flá ‘then’).
Phrases or clauses that have adverbial function are known as
adverbials. But since adverbs by definition also have adverbial func-
tion, they are clearly adverbials too. The difference between the two is
a matter of perspective. Adverbs are a word class on a par with nouns,
adjectives, verbs, etc., while adverbials are functional elements in the
sentence, comparable to subjects, objects, etc. Although this Grammar
attempts as far as practicable to integrate morphology and syntax, its
different sections are rooted firmly in the word class. Adverbials will
therefore not be further discussed. In fact, Old Norse and English do
not differ greatly with respect to adverbial formation and use, so it is
unlikely the student will encounter much difficulty in recognising Old
Norse adverbials for what they are.
Verb inflexions and their function 131
3.6.2 Tense
3.6.3 Mood
tudes to the truth value of what is said on the part of speaker or writer.
Was, as also goes in the previous pair of examples, represents what is
known as the indicative mood, were, together with go in the previous
pair, the subjunctive.
Beyond these cases (3rd person singular present of most verbs and
1st (and 3rd) singular past of to be), there is little of mood inflexion in
English. In Old Norse, in contrast, each verb has two full sets of end-
ings. As in English, indicative endings are used by and large in sen-
tences where the predicate denotes something regarded by the speaker/
writer as factual or certain. Subjunctive endings are found chiefly in
sentences where the predicate denotes something regarded by the
speaker/writer as hypothetical: a wish, request, instruction, supposi-
tion, possibility, etc. In addition, Old Norse has an imperative mood
(used for commands — see below), but this only manifests itself in
the 2nd person singular (some have reckoned with 1st and 2nd pl.
imperatives, but these are distinctive usages, not forms). Contrastive
examples, illustrating differences between indicative, subjunctive and
imperative forms in the present tense and indicative and subjunctive
forms in the past (there is no past imperative) are: flú kastar ‘you
throw’ (indic.) — nema flú kastir ‘unless you throw’ (subj.) — kasta
(flú) (or: kasta›u, cf. 3.2.1) ‘throw!’ (imp.); fleir dœm›u ‘they judged’
(indic.) — fló fleir dœm›i ‘though they judged’ (subj.).
3.6.4 Voice
Voice is a term used to denote the way in which the relationship be-
tween the subject and the object of a verb is expressed. The main
distinction in many languages, and the only one it is useful to make in
Old Norse, is between active and passive. If in English we say John
hit him, the subject is also the agent (i.e. John is the one who does the
hitting, cf. 3.1.5, sentence 1). But we may turn the sentence round, as
it were, and say he was hit (by John), where the subject, ‘he’, is the
goal of the action, or the ‘patient’, and the agent appears (optionally)
in a preposition phrase (cf. 3.7). The first type of construction is known
as active and the second as passive.
Old Norse forms passive verb phrases in much the same way as
English (e.g. fleir halda hátí› mikla ‘they hold a great festival’ (act.)
— hátí› mikil er haldin ‘a great festival is held’ (pass.); fleir nefndu
136 Morphology and syntax
hann Óláf ‘they named him Óláfr’ (act.) — hann var nefndr Óláfr ‘he
was named Óláfr’ (pass.)). Morphologically, such passives are peri-
phrastic: the main verb does not itself inflect for voice, rather we have
the appropriate form of the present or past tense of the verb vera ‘[to]
be’ used as an auxiliary in combination with the past participle of the
main verb, which inflects as an adjective (haldin, nom. f. sg., agrees
with hátí›, nefndr, nom. m. sg., agrees with hann, cf. 3.3.1). (See further
3.9.7.2.)
Old Norse does, however, possess an inflexion that sometimes has
passive function. This is the -sk form of the verb (so-called because in
most manifestations it consists of an -sk added to the appropriate verbal
ending). It would be misleading, however, to consider the -sk a passive
form since it more commonly appears with other functions (see 3.6.5.3).
Examples of passive usage are: hann fyrirdœmisk af illum mƒnnum
‘he is condemned by wicked men’, á hans dƒgum bygg›isk Ísland ‘in
his days Iceland was settled’, hann fannsk eigi ‘he was not found’ ‘he
could not be found’. Fyrirdœmisk ‘is condemned’ contrasts with
fyrirdœmir ‘condemns’, bygg›isk ‘was settled’ with bygg›i ‘settled’
and fannsk ‘was found’ with fann ‘found’.
1. In what ways do person and number affect the form of the verb in
Old Norse?
2. How many tenses may the Old Norse verb be said to have, and
why?
3. How is the past tense distinguished from the present in Old Norse?
4. How is the subjunctive mood marked in Old Norse, and what are
its chief functions?
5. To what extent can the passive voice be expressed by inflexions in
Old Norse?
Verb inflexions and their function 137
3.6.5.1 Endings
This section sets out the inflexions that are attached to the verbal root,
i.e. the personal endings and the past tense suffix of weak verbs (see
3.6.2). Because strong verbs lack a past tense suffix and some of their
personal forms are different from those of the weak verbs, it is clearest
to provide separate tables for the two types. The student should ob-
serve, however, that the majority of personal endings are common to
both strong and weak verbs. With either type, each tense potentially
has six different endings for the indicative and six for the subjunctive
(three persons, 1st, 2nd and 3rd, times two numbers, singular and
plural), and also the 2nd sg. present imperative. In reality the number
is smaller because the same form can occur in more than one position.
The endings are as follows (~ = zero, i.e. there is no ending, the form
consisting of root alone (e.g. ek tek ‘I take’, hon tók ‘she took’); actual
paradigms are given in 3.6.10).
138 Morphology and syntax
Strong verbs
Indicative Subjunctive
Weak verbs
Indicative Subjunctive
and there is therefore nothing like the u of labial mutation or even the
historical i of front mutation to warn us what vowels to expect and
when to expect them. The alternations concerned are not arbitrary,
however, but conform to regular patterns, so as soon as one particular
form of a strong verb is encountered, it is often possible to predict
what the root vowels of all the other forms will be.
Here we are concerned with the present and past tenses. In these a
maximum of three different gradation vowels are found, one through-
out the present (subject to front mutation in the singular indicative
and to breaking in the plural indicative and subjunctive), another in
the singular past indicative, a third in the plural past indicative and the
past subjunctive (the latter also subject to front mutation). In all, there
are six regular gradation series, that is, ways in which root vowels may
alternate, and a few minor patterns found only in a small number of
verbs, albeit some quite common ones. One series, for example, has í
in the present tense, ei in the sg. past indic., and i in the remaining past
tense forms. If therefore we come across the sentence hann greip sver›
sitt ‘he grasped his sword’, we may deduce (a) that greip is a singular
past tense form (in the absence of the -r 3rd sg. present ending or the -i
of the subjunctive, and noting that there is in any case no verb *greipa),
and (b) that the root of the present tense will be gríp- and of the past
plural and past subjunctive grip-. Another series has a in the present, ó
throughout the past. An unfamiliar verb-form fór (there being no *fóra)
may therefore be taken as singular past and its present root confi-
dently assumed to be far-, but with front mutation in the singular present
indicative (cf. hon ferr ‘she goes’, flér fari› ‘you [pl.] go’, hon fór ‘she
went’). The six basic vowel gradation series have the following alter-
nations in the present, past sg. indic., and past pl. indic./past subj.
(front mutation forms are given in brackets):
(1) í —— ei —— i
(2) jó/jú (‡) —— au —— u (y)
(3) e —— a —— u (y)
(4) e —— a —— á (æ)
(5) e —— a —— á (æ)
(6) a (e) —— ó —— ó (œ)
It will be observed that (4) and (5) are identical. This is because a
complete series also takes in the past participle (see 3.6.6), and there
Verb inflexions and their function 143
the root vowel of (4) and (5) does vary. As noted above, certain verbs
which form their past tense by vowel change follow patterns other
than the six just described. We find a — e —e, á — é — é, au — jó —
jó, ei — é — é and variations on each. Most of the few verbs involved
are very common, and it is probably sensible for the student to learn
them individually as they are encountered.
In order to flesh out this rather abstract account, a verb illustrating
each of the six series and the minor patterns is now provided; the forms
are cited in the following order (the pl. past subj. has the same root
vowel as the sg., and indeed the 3rd pl. has exactly the same form as
the 3rd sg.; only the basic meaning(s) of the verb are given):
3rd sg., pl. present indic., 3rd sg., pl. past indic., 3rd sg./pl. past subj.
As indicated above (3.6.4), the -sk form of the verb consists for the
most part of an -sk suffix added to existing endings. Where the final
sound in an ending is -r, this is assimilated to the s and the resulting ss
is then simplified (e.g. finnsk ‘is found’, 3rd sg. present indic.,
< *finnssk < *finnrsk). Where the final sound of an ending is -› or -t,
the juxtaposition with s is rendered z (e.g. fœ›izk ‘are brought up’, 2nd
pl. present indic., < *fœ›i›sk; cf. 2.1.3). This applies even where › or
t is juxtaposed to s after the assimilation of r as just outlined (e.g. gezk
‘is begotten’, 3rd sg. present indic., < getsk < *getssk < *getrsk).
In older texts the 1st person singular forms deviate from this pattern:
they appear with an -umk ending attached to the plural root of the relevant
tense and mood (e.g. ek kƒllumk ‘I am called’, with labially mutated
root kall-, ek rá›umk frá ‘I refrain from’, with root rá›-, contrast ek
ræ› ‘I advise’). 1st person -sk verbs are not very common at all,
however, especially 1st person singular (and very rarely do they have
passive sense either in the singular or plural, cf. rá›umk above).
In younger texts not only is the 1st person sg. -umk replaced by the
2nd/3rd person sg. form, but a bewildering variety of suffixes is found
as well as or in place of -sk, -umk. The more common are -zk (which
spreads from its original domain, cf. above), -s, -z, -st and -zt; hybrids
such as 1st sg. -umsk, -ums also occur. Ultimately, the -st form re-
places all the others, and is the one used in modern Icelandic, Faroese
and Norwegian nynorsk.
Most normalised texts will use the forms set out in the table below,
but even where that is not the case, or the student is confronted with
an unnormalised text, there should be few problems of recognition.
What needs to be remembered is, first: that -umk, -sk, -zk, -s, -z, -st,
-zt, etc. are variant realisations of a single underlying form and choice
of any particular one does not change the meaning; second: that in
most cases the suffix — whichever is employed — will simply be
attached to the verbal ending (e.g. nefndisk ‘named him/herself’ ‘was
named’ consists of nefn-di-sk: root + 3rd sg. past ending + -sk); the
exceptions to this rule have been described above.
With these reservations, the -sk form of the verb may be set out as
follows. (Both personal and -sk endings are given; actual paradigms
will be found in 3.6.10.)
Verb inflexions and their function 145
Strong verbs
Indicative Subjunctive
Weak verbs
Indicative Subjunctive
So far we have discussed only present and past tense forms of the
verb. The reason for treating these separately, and first, is that they are
Verb inflexions and their function 147
central to every sentence. It was pointed out in 3.6 that the verb is the
‘non-reducible part of the predicate’, but it would be more precise to
say that it is the tensed verb that is the essential element — and in
Germanic languages that means a verb in the present or past tense.
Thus we may attest in English: he sings and he sang, but not: *he sing,
*he sung or *he singing. Nevertheless, sing, sung and singing are con-
sidered to belong to the same lexical item (dictionary word) as sings
and sang, and to that extent to represent the same word class. In terms
of function, however, sing, as in to sing, behaves more like a noun
(compare I want to sing and I want beer, in which to sing and the noun
beer occupy the same slot in the sentence), and sung and singing more
like adjectives (compare a sung chorus, the singing detective and a
noisy chorus, the smart detective, in which sung, singing occupy the
same slots as the adjectives noisy, smart; note that singing may also
be a pure noun as in I like singing, but then it is not considered part of
the verb at all).
There is thus every reason to make a distinction between to sing,
sung and singing on the one hand and sings and sang on the other. In
grammatical description the former are commonly said to represent
the non-finite parts of the verb, the latter the finite. This terminology
is based on the observation that sings and sang make a contrast of
tense; they are in one way or another bound by time. The same is not
true of to sing, sung and singing, which are independent of time. That
is perhaps not immediately obvious in the case of sung or singing.
Sung appears to refer to the past (I have sung mass), and is even called
a ‘past participle’. Consider, however, the hymn was/is/will be sung in
unison, where the time distinctions are not applicable to sung, but are
in the finite verbs, was/is/will. Singing is even harder to connect with
past, present or future. It is known as a ‘present participle’, but is in
fact timeless (cf. the singing detective); in verb phrases of the type
was/is/will be singing, it is again the finite verbs that provide the time
reference.
Old Norse has the same non-finite forms as English, to wit: the
infinitive — at syngja ‘to sing’, the past participle — sunginn ‘sung’,
and the present participle — syngjandi ‘singing’. Mention is occa-
sionally made of a ‘past infinitive’, but the form concerned is in origin
the 3rd pl. past indic. and its use as an ‘infinitive’ seems to have arisen
through the recasting of certain finite clauses on analogy with com-
mon constructions that employ the standard infinitive. Very few ‘past
148 Morphology and syntax
infinitive’ forms are attested, in Old Norse prose only three regularly.
The usage is illustrated in 3.9.4.
Being non-finite forms, the infinitive and the participles do not have
verbal inflexion. The Old Norse infinitive is not inflected at all. It
regularly ends in -a, to which the -sk form may be suffixed as appro-
priate (e.g., berja ‘beat’, berjask ‘fight [literally: beat each other]’).
The participles, as we have seen (3.3.9, paradigms 7, 11, 16, 19; also
3.3.6, sentences 1, 4, 7, 22), inflect as adjectives. The past participle
of strong verbs has the adjectival -in suffix, that of weak the same -›,
-d, -t suffix as the past tense (e.g. farinn, farit ‘gone’, from fara, kraf›r,
kraft ‘demanded’, from krefja, strong nom. m. and nom./acc. n. sg. in
both cases). The -sk inflexion is added to the nom./acc. n. sg. form in
various periphrastic constructions (e.g. hafa farizk ‘have perished’ (farit
+ sk, with ts written z), var sætzk ‘was come to terms [i.e. terms were
agreed]’ (sætt + sk)). The present participle is formed with an -and
suffix, as shown in 3.3.9, paradigm 19. It does not normally take the
-sk inflexion.
Although the non-finite verb-forms in terms both of inflexion and
function are largely non-verbal, they are nevertheless, as noted above,
considered to belong to the same word class as the finite. This is be-
cause it is counter-intuitive to view the present and past tense of any
given verb as a separate word from the infinitive and the participles.
The non-finite forms thus have their place in the verbal paradigm.
Indeed, the infinitive is usually taken as the basic form — as the word
itself, of which all the other manifestations are inflected parts. That is
why the infinitive regularly appears as the dictionary entry form.
As we have seen, the endings of verbs in Old Norse and the root
vowel alternations caused by labial and front mutation and breaking
are predictable. This means that it is only necessary to cite a minimal
number of basic forms for the student to be able to identify a particular
verb-form s/he has encountered, i.e. to determine what verb it is part
of and its person, number, tense and mood. These basic forms, known as
‘principal parts’, include the infinitive and the past participle. From
the infinitive it is possible to deduce all the present tense forms (pro-
vided the person and number endings and the workings of labial and
front mutation and breaking are known). From the 3rd sg. past indicative
(or alternatively the 1st or 2nd person) all the past tense forms of weak
verbs can be readily predicted. This is less true of strong verbs: many
Verb inflexions and their function 149
undergo vowel change between the singular and plural past indicative,
so they need to be cited in both a singular and plural form; the past
subjunctive of strong verbs, on the other hand, can be deduced from
the plural indicative (once again, provided the inflexional basics are
known). Finally, the past participle needs to be given since those of
strong verbs usually exhibit further root vowel change; a few weak
verbs, too, show irregular forms, but for the most part their participles
are deducible from the past tense. We thus have a minimum of three
principal parts for weak verbs and four for strong. Front-mutated
present singular indicatives and past subjunctives may be included as
optional extras, but these are non-essential. The decision whether or not
to cite them will depend on how much help one thinks the learner needs.
This is how the system works. A strong verb like rjúfa ‘break’ ‘vio-
late’ will be listed in a grammar or dictionary with its infinitive, rjúfa,
the 3rd (or 1st) sg. past indic. rauf, 3rd (or 1st) pl. past indic. rufu
(rufum), and pp. rofinn or rofit (the choice in the case of the pp. being
between the strong nom. m. or nom./acc. n. sg. forms). From the
infinitive, the present sg. indic. forms r‡f, r‡fr can be deduced by
applying the appropriate endings and the rule: ‘in the present sg. indic.
strong verbs with back root vowels undergo front mutation’. All the
other present tense forms will have root rjúf-. The 1st, 2nd and 3rd sg.
past indic. have root rauf- with the -t ending added in the 2nd person.
The 1st, 2nd and 3rd pl. past indic. have root ruf- plus the appropriate
endings, and from this the subjunctive root ryf- can be deduced, to
which the appropriate subjunctive endings are added. The participial
root is rof-, which remains unchanged whatever the adjective ending.
A weak verb like krefja ‘demand’ will be listed with infinitive krefja,
3rd (or 1st) sg. past indic. kraf›i (kraf›a), and pp. kraf›r or kraft. From
the infinitive all the present tense forms can be deduced simply by
adding the appropriate endings. (One will need to recognise the type
of weak verb involved (see below) to know whether the indicative sg.
endings are ~, -a or -i (1st), -r, -ar or -ir (2nd/3rd), but this variation
is unlikely to cause problems of understanding to the reader of Old Norse.)
From the 3rd or 1st sg. past indic. all the past tense forms can be
deduced by applying the appropriate endings, the labial mutation rule,
and the rule: ‘all disyllabic past subjunctive forms with original back
root vowels exhibit front mutation’. Thus, the pl. indic. root + -› suffix
of krefja will be krƒf›-, because all three plural endings begin with u;
150 Morphology and syntax
The three types of weak verb differ in a number of ways. For the
learner what will be most noticeable is: type 1 has root vowel change
between the present and past indic. (krefja — kraf›i) and no vowel in
the sg. present indic. endings (hann krefr); type 2 has a ‘connecting
vowel’ a in the past tense (kalla›i) and a in the sg. present indic. end-
ings (hann kallar); type 3 has the same root vowel throughout, no
connecting vowel in the past tense and i in the sg. present indic. end-
ings (hann heyrir). The three distinct past tense suffixes, -›, -d and -t,
are distributed not according to type of verb, but phonetic environ-
ment, so that › occurs after vowels and most voiced consonants (kalla›i,
fá›i ‘coloured’, kraf›i, heyr›i), d chiefly after n (hefndi ‘avenged’),
and t after unvoiced consonants (vakti ‘wakened’, œpti ‘shouted’). In
the earliest texts fl is found after unvoiced consonants, and from the
late thirteenth century onwards d replaces › after certain voiced
consonants, particularly l and m (val›i/valdi ‘chose’, dœm›i/dœmdi
‘judged’).
152 Morphology and syntax
The preterite present verbs of Old Norse form a small but important
class — important because virtually all its members are extremely
common. The majority are modal auxiliaries (verbs subordinate to the
main verb, which express mood, e.g. English I would come, she might
go; cf. 3.6.3). The term ‘preterite present’ reflects the fact that verbs of
this type have strong past tense forms in the present; in the past they
inflect for the most part like weak verbs, though not all of them have
the dental suffix associated with weak inflexion. The reason for the
preterite present aberration lies in linguistic pre-history. Put at its most
simple, the Germanic past tense is a development of an earlier perfect,
which expressed completed action or the state obtaining after the action.
While the perfects of most verbs happily made the transition to past,
those of what became the preterite presents seem so firmly to have
expressed present state that they were ultimately absorbed into the
present tense by the creation of new (weak) past tense forms. Thus,
ON vita ‘know’ is related to Latin videre ‘see’ ‘perceive’, and hon veit
‘she knows’ (cf. past tense beit ‘bit’, leit ‘looked’, etc.) must derive
from a form that originally meant something like ‘she has perceived’.
Verb inflexions and their function 153
The principal parts of the preterite presents are listed in the follow-
ing order (the pp. is given in the strong nom./acc. n. sg. form, for some
verbs the only one used; note the infinitives munu and skulu, mod-
elled on the 3rd pl. present indic., which in virtually all verbs has the
same form as the infinitive):
inf., 3rd sg., pl. pres. indic., 3rd sg. pres. subj., 3rd sg. past indic., subj., pp.
eiga ‘own’
eiga —— á —— eigu —— eigi —— átti —— ætti —— átt
muna ‘remember’
muna — man — munu — muni — mundi — myndi — munat
unna ‘love’
unna — ann — unnu — unni — unni — ynni — unnt/unnat
vita ‘know’
vita — veit — vitu — viti — vissi — vissi — vitat
flurfa ‘need’
flurfa — flarf — flurfu — flurfi — flurfti — flyrfti — flurft/flurfat
preterite presents, and vera ‘be’, a highly irregular strong verb basically
of type 5 but with preterite-type forms in the present indic. (2nd sg.
ert, 2nd, 3rd pl. eru›, eru; note also 1st sg. em). Of vilja the same
principal parts are given as for the preterite presents above; of vera
the same plus the 3rd pl. past indic. (cf. the principal parts of strong
verbs in 3.6.6 above):
Following the same procedure as for other word classes, examples are
now given of verbs in function. With the vast range of verbal forms
and functions that exists, only a selection can be illustrated, with the
emphasis on the most common types. Equally, because so many dif-
ferent features are involved — person and number, tense, mood, voice,
-sk forms, periphrastic constructions — and several features combine
in the one verb phrase, it has proved difficult to order the examples in
any meaningful way. Note that the verbal inflexions being illustrated
(or the whole word where there is no difference from the root of the
infinitive or an inflexion cannot easily be discerned) are printed in
bold type. To underline the grammatical relations involved, bold is
also used for the subject, which triggers the person and number form
in the verb. Compare the inflexions used below with those set out and
discussed in 3.6.5, 3.6.6 and 3.6.7.
B‡r is 3rd sg. present indic. of the strong verb búa (minor type). Fór is 3rd sg.
past indic. of the strong verb fara (type 6). Indicative is used because factual
statements are being made about what happened. The abrupt change from
present to past tense is characteristic of Old Norse prose style.
156 Morphology and syntax
(2) Jarl svarar ok ba› konung gefa sér frest at hugsa fletta mál
‘Earl answers and bade king give self respite to consider this
matter’
‘The earl answers and asked the king to give him time to
consider this matter’
Svarar is 3rd sg. present indic. of the weak verb svara (type 2). Ba› is 3rd sg.
past indic. of the strong verb bi›ja (type 5, but with root i in the inf. and
present tense, see 3.6.9.1 point (5)). Indicative is used in both cases because
factual statements are being made about what happened. Gefa is an infinitive,
a complement of ba›; it has no overt subject, but konung, the object of ba›,
functions as covert (understood) subject (i.e. it is the king who is to do the
giving; see further (24) below and 3.9.4). Hugsa is likewise an infinitive, a
complement of frest; again there is only a covert subject: the earl (i.e. it is he
who is to do the considering).
(3) fiorfinnr vissi eigi, at Brúsi haf›i upp gefit ríki sitt
‘fiorfinnr knew not that Brúsi had up given realm REFL. POSS.’
‘fiorfinnr did not know that Brúsi had surrendered his realm’
Vissi is 3rd sg. past indic. of preterite present vita. Haf›i is 3rd sg. past indic.
of weak hafa (type 3, but irregular, see 3.6.7); together with the pp. gefit, from
gefa (strong type 5), it forms a so-called ‘past perfect’ construction, the equiva-
lent of English ‘had given’ (the strong nom./acc. n. sg. form of the pp., when
used in perfect and past perfect constructions, is known as the supine, see
3.9.7.1). On the use of the indicative mood, see (1) and (2) above.
Skil›usk is 3rd pl. past indic. of weak skilja (type 1) with the -sk suffix (skil›u
+ sk). On the use of the indicative, see (1) and (2) above. Skilja means ‘sepa-
rate’ ‘divide’; the -sk form imparts a reciprocal sense: ‘they separated (from)
each other’.
Sefask is 3rd sg. present indic. of weak sefa (type 2) with the -sk suffix (sefar
+ sk with assimilation rs > ss and simplification ss > s in unstressed position
(see 3.6.5.3)). On the use of the indicative, see (1) and (2). Sefa means ‘soothe’
Verb inflexions and their function 157
‘calm’; the -sk form is probably in origin a reflexive (‘calms himself’), but it
can also be conceived as passive (‘is soothed’), and thus illustrates how the
function of the -sk form could develop from reflexive to passive.
Segja is 3rd pl. present indic. of weak segja (type 3, but with vowel change in
the past tense, see 3.6.9.2 point (5)). Hafi is 3rd sg. present subj. of weak hafa
(see (3)); together with supine fallit, from falla (strong minor type), it forms a
perfect construction (see (3)). Observe the difference between the use of the
indicative and subjunctive: that ‘men say’ is what the writer reports as fact;
that ‘he has fallen’ is not what the writer says, but what he claims other people
say, and thus from the writer’s point of view no longer a statement of fact.
Vill is 2nd sg. present indic. of weak vilja (type 1, but irregular, see 3.6.7 and
3.6.9.1 point (11)); together with inf. gerask, -sk form of weak gera ‘do’ ‘make’
(type 3, but irregular, see 3.6.7), it forms a modal construction (see 3.6.3).
Gerask has a different meaning from gera, though the origin of the sense
‘become’ can probably be sought in the reflexive ‘make oneself’. Er is 3rd sg.
present indic. of irregular vera (3.6.7). Setja is 1st sg. present subj. of weak
setja (type 1, but with no vowel alternation between present and past, see
3.6.9.3). Vil is 1st sg. present indic. of vilja (see above); here, too, it functions
as a modal, although not accompanied by an overt infinitive (the sense, how-
ever, is ‘whom I want to put’). Observe the difference between the use of the
indicative and subjunctive. In present tense conditional sentences beginning
with ef ‘if’ (see 3.8.2.4) the indicative is normally used even though no state-
ment of fact is being made, hence vill. In the independent sentence which
follows there is hardly a recording of fact either, rather a statement of the
situation that will obtain if the condition is not fulfilled, but such sentences,
too, have the indicative. Setja, however, denotes a wholly hypothetical action,
and is accordingly subjunctive. With vil we are once again back with the
indicative: the speaker’s will is presented as real and immediate.
158 Morphology and syntax
Tók is 3rd sg. past indic. of strong taka (type 6, but with root e in the pp., see
3.6.9.1 point (4)). Gneri is 3rd sg. past indic. of irregular gnúa (3.6.7). Both
statements are factual and the indicative is therefore used.
Váru is 3rd pl. past indic. of irregular vera (3.6.7); together with the pp.s
drepnir and reknir, from drepa and reka (both strong type 5), it forms passive
constructions, the equivalent of English ‘were killed’, ‘(were) driven’ (in such
constructions the pp. inflects as a strong adj. (see 3.6.4), here nom. m. pl.,
agreeing with the subjects sumir . . . sumir). For the use of the indicative, see
(1) and (2).
Hefir is 2nd sg. present indic. of weak hafa (type 3, but irregular, see 3.6.7);
together with the supine heyrt, from heyra (weak type 3), it forms a perfect
construction (see (3)). Em is 1st sg. present indic. of irregular vera (3.6.7).
Bœta is an infinitive, a complement of vera vanr ‘be accustomed’; its covert
subject is the ek of the finite sentence: ek em ekki vanr (see (2)). Læt is 1st sg.
present indic. of strong láta (minor type); it acts here as an auxiliary, and
together with the infinitive (drepa, strong type 5) forms a construction with
the sense ‘cause to be killed’ ‘have killed’. Indicative is used throughout be-
cause nothing is presented as unreal or hypothetical; after verbs meaning ‘hear’,
‘learn’, ‘discover’, etc., the indicative is almost always found, the truth value
of what is ‘heard’ being taken for granted; the unwillingness of the speaker to
pay compensation and his propensity to have people killed are in no doubt.
Herju›u is 3rd pl. past indic. of weak herja (type 2). Kómu, fóru and námu are
likewise 3rd pl. past indic., of strong koma (historically type 4, but highly
irregular, see 3.6.9.3), fara (type 6) and nema (type 4). On the use of the
indicative, see (1) and (2).
(12) Fyrir ofdrambs sakar haf›i hann villzk ok snúizk ífrá gu›i
‘For arrogance sake had he bewildered-sk and turned-sk from
God’
‘Because of arrogance he had gone astray and turned from God’
Haf›i is 3rd sg. past indic. of weak hafa (type 3, but irregular, see 3.6.7);
together with the -sk supines villzk and snúizk (< villt + sk, from weak type 3
villa, snúit + sk, from irregular snúa (3.6.7), both with ts written ‘z’) it forms
past perfect constructions (see (3)). Both the -sk forms are in origin probably
reflexives (‘led himself astray’, ‘turned himself’). On the use of the indica-
tive, see (1) and (2).
(13) Muntu ok eigi vilja vita flat á flik, at flú liggir hér sem kƒttr
í hreysi, flar er ek berjumk til frelsis hvárumtveggjum
‘Will-you also not want know that onto you, that you lie
here like cat in cranny, there where I fight-sk for freedom
for-both’
‘You will also not want to be accused of lying here like a cat
in a cranny while I fight for the freedom of both of us’
Muntu (either munt + flú with assimilation tfl > tt and simplification tt > t after
another consonant or mun + flú with loss of -t ending before flú and partial
assimilation nfl > nt, see 3.2.1) is 2nd sg. present indic. of preterite present
munu; together with infinitives vilja (weak type 1, but irregular, see 3.6.7,
3.6.9.1 point (11)) and vita (preterite present) it forms a double modal con-
struction (i.e. two modal verbs ‘will [future]’ and ‘want to’ are involved).
Liggir is 2nd sg. present subj. of strong liggja (type 5, but irregular, see 3.6.9.3).
Berjumk is 1st sg. present of weak berja (type 1) with the -umk suffix (which
160 Morphology and syntax
replaces -sk in the 1st sg.). The -sk form of berja is in origin reciprocal (‘beat
each other’), but it comes to have the more general meaning ‘fight’ — in
which ‘each other’ may or may not be understood. Of the three finite verbs in
this example one is indic., one subj. and one indeterminate on the basis of
form: mun(t) records what the speaker presents as fact, whereas liggir refers
to a hypothetical event; berjumk is almost certainly indic. since the speaker is
in no doubt about the fighting in which he will be involved.
Veit is 3rd sg. present indic. of preterite present vita. Ver›a is 1st sg. present
subj. of strong ver›a (type 3, but irregular, see 3.6.7). The first sentence con-
tains a direct present-tense question introduced by an interrogative pronoun
(hverr) and, like all sentences of this type, has a verb in the indicative. The
second sentence is introduced by the conjunction nema which automatically
triggers a subjunctive verb-form since it presupposes a hypothetical situation.
Bei›ir is 3rd sg. present indic. of weak bei›a (type 3). Skyli is 3rd sg. present
subj. of preterite present skulu; together with inf. rá›ask, -sk form of strong
rá›a ‘advise’, ‘rule’ (minor type), it forms a modal construction (see 3.6.3).
Rá›ask has various meanings, mostly different from those of rá›a; the semantic
development can often be hard to trace. Bei›ir is indic. because it denotes
what the writer regards as fact; skyli, in contrast, refers to what Einarr wants
to happen, but which may or may not take place.
Hru›usk is 3rd pl. past indic. of hrjó›a (strong type 2) with the -sk suffix. The
sense of -sk here is clearly passive: some agency cleared the ships (i.e. killed
those on board) but the goal of the action, ‘ships’, has been made subject and
the agent is left unexpressed. On the use of the indicative, see (1) and (2).
Verb inflexions and their function 161
Var› is 3rd sg. past indic. of strong ver›a (type 3, but irregular, see 3.6.7);
together with gƒr, pp. of gøra/gera (weak type 3, but irregular, see 3.6.7), it
forms a passive construction (see (9)). Ver›a, as well as vera, may be used as
the equivalent of English ‘be’ in passive verb phrases (see further 3.9.7.2). On
the use of the indicative, see (1) and (2).
Ver›ir is 2nd sg. present subj. of strong ver›a (type 3, see 3.6.7). The conjunc-
tion fló at or flótt (3.8.2.2), which introduces the first sentence, automatically
triggers a subjunctive verb-form since it mostly presupposes a hypothetical
situation. Mældu (mæl + flú, with partial assimilation lfl > ld, see 3.2.1) is the
imperative of mæla (weak type 3) with the subject pronoun attached; it ex-
presses an instruction.
Vil is 1st sg. present indic. of weak vilja (type 1, but irregular, see 3.6.7, 3.6.9.1
point (11)). As a modal auxiliary, it is regularly followed by an inf., but here
that is replaced by the dependent sentence at flit hittizk. Hittizk is 2nd pl. present
of hitta (weak type 3) with the -sk suffix (›s being written ‘z’). The sense of -sk
here is reciprocal: ‘meet each other’. The mood of the verb cannot be deduced
from the form, but it is almost certainly subj., determined by the sense of the pre-
ceding independent sentence: that which is wanted or wished for is hypothetical.
Ætla›a is 1st sg. past indic. of weak ætla (type 2). Mynda is 1st sg. past subj.
of preterite present munu; together with infinitives koma (strong type 4 his-
torically, but highly irregular, see 3.6.9.3) and gjalda (strong type 3, see 3.6.5.2)
it forms modal constructions (3.6.3). Væra is 1st sg. past subj. of irregular
vera (3.6.7). Geld is 1st sg. present indic. of strong gjalda (see above). Hefi is
1st sg. present indic. of weak hafa (type 3, but irregular, see 3.6.7); together
with supine gefit, from gefa (strong type 5), it forms a perfect construction
(see (3)). The three subjunctives, mynda (twice) and væra, all depend on ætla›a
in the independent sentence: this is what the speaker thought would happen,
but events have proved him wrong. With geld, we are back to statements the
speaker presents as factual.
(21) Hann veitti allri hir› sinni bæ›i mat ok mungát, svá at menn
flyrfti eigi í skytning at ganga
‘He gave all his retainers both food and ale, so that men would
not need to go to an inn’
Veitti is 3rd sg. past indic. of weak veita (type 3). fiyrfti is 3rd pl. past subj. of
preterite present flurfa; together with inf. ganga (strong minor type) it forms a
modal construction (see 3.6.3). Indic. veitti is used in what the writer presents
as a statement of fact. The subjunctive flyrfti suggests a purpose rather than a
result sentence: svá at ‘so that’ can mean either ‘in order that’ or ‘with the
result that’ (see 3.8.2.2); the former is putative, normally requiring the subj.,
the latter factual, normally requiring the indic.
(22) Ef hann væri heill at sumri, sag›i hann, at fleir skyldi finnask
‘If he were hale at summer, said he, that they should find-sk’
‘If he were alive when summer came, he said, they should meet’
Væri is 3rd sg. past subj. of irregular vera (3.6.7). Sag›i is 3rd sg. past indic.
of weak segja (type 3, but with vowel change in the past tense, see 3.6.9.2
point (5)). Skyldi is 3rd pl. past of preterite present skulu; together with inf.
finnask, -sk form of strong finna ‘find’ (type 3, but irregular, see 3.6.9.2 point
(2), 3.6.9.3), it forms a modal construction (see 3.6.3). The -sk form has recip-
rocal sense: ‘find each other’, and thus ‘meet’. Indicative sag›i presents what
the writer regards as fact, namely that ‘he’ said the accompanying sentences.
Væri conforms to the usage whereby past tense verbs in conditional sentences
are almost always subjunctive (even when, as here, the condition is ‘open’,
i.e. may or may not be fulfilled, and the past tense form is simply the reported
speech equivalent of direct: ‘if I am alive when summer comes’). The mood
of skyldi cannot be deduced from the form, but it is certainly subj., referring to
hypothetical circumstances dependent on the indirect-speech condition of ef
hann væri heill at sumri.
Verb inflexions and their function 163
(23) Mun samflykki okkart mest, at vit innimsk lítt til um flann
flri›jung landa
‘Will concord our [dual] greatest, that we allude-sk little to
about that third of-lands’
‘Our concord will be greatest if we make little mention of
that third of the country’
Mun is 3rd sg. present indic. of preterite present munu; vera ‘be’, with which
it forms a modal construction, is omitted but understood (see 3.9.5.2). Innimsk
is 1st pl. present subj. of weak inna (type 3) with the -sk suffix (innim + sk).
The -sk form is in origin reciprocal: ‘speak to each other’. Indicative mun
expresses what the speaker regards as certain, subjunctive innimsk the hypo-
thetical situation he envisages.
Kalla›i is 3rd sg. past indic. of weak kalla (type 2). Gefit is the pp. of strong
gefa (type 5), acc. n. sg., agreeing with flat ríki; together with verit, supine of
irregular vera (3.6.7), it forms a passive construction (see (9)). Verit for its
part joins with inf. hafa (weak type 3, but irregular, see 3.6.7) to form a per-
fect. We thus have a non-finite perfect passive construction. The lack of a
finite verb arises because the complement of kalla›i is what is known as an
‘accusative and infinitive’ clause — one that takes the object of the matrix
verb as its subject. This is all somewhat complex, so a detailed analysis is now
offered: kalla›i (finite verb), hann (subject), flat ríki (direct object of kalla›i
and subject of gefit), gefit hafa verit (non-finite perfect passive construction),
sér (indirect object of the infinitive clause, but coreferential with the subject
of the independent sentence); a semi-literal translation is: ‘he said that realm
to have been given to himself’. (Some would argue that sér is subject of the
infinitive clause and flat ríki object. These theoretical considerations need not
concern the learner, but see 3.9.3. On acc. + inf. clauses, see further 3.9.4.)
(25) fiér skulu› nú frá mér fless mest njóta, er flér gáfu› mér líf
ok leitu›u› mér slíkrar sœm›ar sem flér máttu›
‘You [pl.] shall now from me that most enjoy, that you [pl.]
gave me life and sought for-me such honour as you [pl.] could’
‘What chiefly benefits you now as far as I am concerned is
that you gave me my life and tried to show me as much hon-
our as you could’
164 Morphology and syntax
Skulu› is 2nd pl. present indic. of preterite present skulu; together with inf.
njóta (strong type 2) it forms a modal construction (see 3.6.3). Gáfu› is 2nd
pl. past indic. of strong gefa (type 5). Leitu›u› is 2nd pl. past indic. of weak
leita (type 2). Máttu› is 2nd pl. past indic. of preterite present mega; although
not accompanied by an overt infinitive, it functions as a modal (the sense is
‘as you could show me’). The indic. is used throughout because everything
said is perceived by the speaker as factual.
Other parts of the verbal system are equally transparent. Those who
have studied the preceding sections will not fail to recognise hann
haf›i drepit tvá menn ‘he had killed two men’ as a past perfect con-
struction and tveir menn váru drepnir ‘two men were killed’ as pas-
sive. The -sk suffix is also hard to confuse with any other ending (though
occasional uncertainty may arise when it appears in its -st, -zt mani-
festations).
Less easy to spot is the difference between indicative and subjunc-
tive mood. To get this right consistently the student will have to be
familiar with the relevant endings, but quite often it is enough to recog-
nise the form of the root (contrast hann drap ‘he killed [indic.]’ with
flótt hann dræpi ‘though he killed [subj.]’). How far it is essential to
know whether a verb-form is indicative or subjunctive will depend on
the context. As the examples in 3.6.8 show, the choice between the
moods is sometimes automatic, sometimes dependent on meaning,
though the differences of meaning can often be subtle and difficult to
render in English.
In the light of these considerations, the deviations from the estab-
lished patterns of verbal inflexion to be concentrated on here are chiefly
those affecting principal parts. The presentation will be divided into
three major sections. First, deviations that follow phonological rules
the student can apply; second, unpredictable deviations that affect a
group of verbs; third, idiosyncratic deviations.
Type 3 binda ‘bind’ (on root vowel i, see 3.6.9.2 point (2)), gjalda
‘pay’, for example, have imp. bitt, gjalt, 3rd sg. past indic. batt, galt;
minor types ganga ‘walk’, halda ‘hold’ have imp. gakk, halt, 3rd sg.
past indic. gekk, helt (sometimes regular imp. forms are encountered
— in the above cases: bind, gjald, gang, hald).
(4) Pp.s of type 6 and minor type strong verbs normally undergo
front mutation of the root vowel when the root ends in -g or -k, e.g.
dreginn from draga, genginn from ganga, tekinn from taka ‘take’.
(5) Present roots of type 6 strong verbs undergo front mutation of
the root vowel when j occurs before endings consisting of or begin-
ning with a or u, e.g. hefja ‘lift’, sverja ‘swear’, 3rd sg. past indic. hóf,
sór (see (1) above). Note also that the same conditions give root vowel
i instead of e in type 5 strong verbs, e.g. bi›ja ‘ask’, sitja ‘sit’.
(6) Pp.s of type 3 and 4 strong verbs have root vowel u rather than o
when the immediately following consonant is m or n, e.g. bundinn
from binda ‘bind’, sprunginn from springa ‘spring’ ‘burst’, unninn
(see (1) above) from vinna ‘work’ (on root vowel i, see 3.6.9.2 point
(2)), numinn from nema ‘take’.
(7) Weak verbs undergo a number of consonantal assimilations and
simplifications when the past tense and participial suffixes -›, -d, -t
are added. Such phonological adjustments are not restricted to verbs,
but are found elsewhere in the language (see 3.1.7.4 point (1), 3.3.8.4
point (2), 3.3.8.5 point (2)). Verbs whose root ends in consonant + ›, d
or t do not add a further ›, d or t to mark the past-tense or participial/
supine suffix, e.g. vir›a — vir›i — vir›r ‘value’, senda — sendi —
sendr ‘send’, svipta — svipti — sviptr ‘deprive’. This applies equally
when the root ends in tt, e.g. rétta — rétti — réttr ‘straighten’ ‘stretch
out’. Verbs whose root vowel is immediately followed by › show
assimilation ›d > dd in the past tense and past participle, e.g. ey›a
— eyddi — eyddr ‘destroy’, gle›ja — gladdi — gladdr ‘gladden’. The -t
ending of the nom./acc. n. sg. of the pp. regularly amalgamates with
the participial suffix (by processes of simplification or assimilation
and simplification; see further 3.3.8.4 point (2) and 3.3.8.5 point (2)),
e.g. flutt (< flutt + t) from flytja ‘convey’, kastat (< kasta› + t) from
kasta ‘throw’, sent (< send + t) from senda, leyst (< leyst + t) from
leysa ‘loosen’, ‘resolve’, hitt (< hitt + t) from hitta ‘meet’.
(8) As with nouns and adjectives (3.1.7.5 point (2), 3.3.8.5 point
(4)), the vowels of endings tend to be dropped when they immediately
Verb inflexions and their function 167
follow a long vowel of the same or similar quality. Thus weak type 3
trúa ‘believe’, for example, has a 1st pl. present indic. form trúm
(< *trúum)’, deyja ‘die’ 3rd pl. past indic. dó (< *dóu), fá ‘get’ 1st pl.
present indic. fám (< *fáum), sjá ‘see’ pp. sénn (< *séinn) (these last three
verbs are highly irregular and their principal parts are listed in 3.6.9.3).
(9) As with adjectives, t is lengthened when immediately following
long, stressed vowels. Thus the 2nd sg. past indic. of strong type 1
stíga ‘step’ (see (2) above) is stétt, of búa ‘prepare’, ‘dwell’ (3.6.9.3)
bjótt.
(10) Strong verbs whose root ends in -› or -t suffer changes to these
consonants in the 2nd sg. past indic. The › assimilates to the -t ending
(cf. (7) above), e.g. reitt, from strong type 1 rí›a ‘ride’. Where the root
ends in -t, the usual ending is -zt, e.g. bazt, from strong type 3 binda
‘bind’ (3rd sg. past indic. batt, see (3) above; on root vowel i see
3.6.9.2 point (2)), lézt, from strong minor type láta ‘let’. This latter
change affects preterite present vita ‘know’ too (2nd sg. present indic.
veizt). Some verbs with root final -› may have the -zt ending as an
alternative to -tt, e.g. bazt or batt from strong type 5 bi›ja (on root
vowel i, see 3.6.9.1 point (5)). Some with root final -t may as an
alternative add t in the normal way, e.g. létt from láta (see above), or
have the same form as the 1st and 3rd sg. past indic., e.g. helt from
strong minor type halda ‘hold’ (see (3) above). Strong verbs with root
final -st have zero ending in the 2nd sg. past indic., e.g. laust from
strong type 2 ljósta ‘strike’.
(11) As in the case of nouns and adjectives (3.1.7.4 point (1), 3.3.8.4
point (1)), an -r ending may sometimes be assimilated to an immedi-
ately preceding l, n or s, e.g. 3rd sg. present indic. vill (< *vilr), from
irregular weak type 1 (3.6.7) vilja ‘want’, skínn (< *skínr) from strong
type 1 skína ‘shine’, les(s) (< *lesr) from strong type 5 lesa ‘gather’, ‘read’.
(12) The 2nd sg. past indic. -t ending of strong verbs is often dropped
when the 2nd person pronoun immediately follows, e.g. gekkt flú or
gekk flú ‘you went’, tókt flú or tók flú ‘you took’.
(13) The 1st pl. -m ending is often dropped when the 1st person dual
or pl. pronoun immediately follows, e.g. tƒkum vit or tƒku vit ‘we two
take’, tókum vér or tóku vér ‘we took’.
(14) The 2nd pl. -› ending is often dropped when the 2nd person
dual or pl. pronoun in the form flit, flér immediately follows, e.g. taki›
flit or taki flit ‘you two take’, tóku› flér or tóku flér ‘you took’.
168 Morphology and syntax
(15) The 3rd sg. present indic. -r ending of the verb flyk(k)ja ‘seem’
is often dropped when the dat. of the 1st or 2nd person sg. pronoun
immediately follows, e.g. flyk(k)ir mér or flyk(k)i mér ‘it seems to me’.
(1) A few strong verbs of type 2 have present tense root vowel ú
rather than jó or jú, e.g. lúta ‘bend down’, súpa ‘sip’.
(2) Several strong verbs of type 3 have present tense root vowel i
rather than e or ja, and a few have y or ø, e.g. binda ‘bind’, finna
‘find’, syngva/syngja (see (6) below) ‘sing’, søkkva ‘sink’. The verbs
with present tense i and y have root vowel u in the pp. (see 3.6.9.1
point (6)); those with present y and ø have root vowel ƒ in the past sg.
indic., e.g. sƒng ‘sang’, sƒkk ‘sank’.
(3) Some weak verbs of type 1 and type 3 have pp.s like those of
type 2, e.g. huga›r (or hug›r) from hyggja ‘think’ ‘intend’, viljat from
vilja ‘want’, florat from flora ‘dare’. Many type 1 verbs have alterna-
tive pp. forms with connecting vowel -i-, e.g. bar›r or bari›r ‘beaten’.
Because the nom./acc. n. sg. of the i-forms is identical with the
nom./acc. n. sg. of the pp. of strong verbs (barit ~ farit), we also get
analogical ‘strong’ pp.s of type 1 weak verbs, e.g. barinn nom. m. sg.
(4) A few weak verbs of type 3 have an -i ending in the imperative
as well as zero, e.g. vak or vaki from vaka ‘keep awake’. The imp. of
flegja ‘stay silent’ is always flegi.
(5) The type 3 weak verbs segja ‘say’ and flegja (3rd sg. pres. indic.
segir, flegir) have root vowel a in the past indic., e in the past subj.,
like type 1 verbs (3rd sg. sag›i, flag›i, seg›i, fleg›i respectively).
(6) As with nouns and adjectives (3.1.7.5 point (4), 3.3.8.5 point (5)), j
may be found in some verbs before endings consisting of or beginning
in a or u; in others v may be found before endings consisting of or
beginning in a or i. With most verbs such insertions are found only in
connection with the present root, but type 2 weak verbs have them through-
out the paradigm. Examples are: svíkja ‘betray’ (strong type 1) — 1st pl.
present indic. svíkjum — 3rd pl. past indic. sviku, syngva ‘sing’ (strong
type 3, on root vowel y, see 3.6.9.2 point (2)) — 3rd pl. present subj.
syngvi — 3rd pl. past subj. syngi, berja ‘strike’ (weak type 1) — 1st sg.
past indic. bar›a, eggja ‘incite’ (weak type 2) — 3rd pl. past indic. eggju›u
Verb inflexions and their function 169
— supine eggjat, bƒlva ‘curse’ (weak type 2) — 3rd pl. past indic.
bƒlvu›u — supine bƒlvat. Note that strong verbs with v insertion and y
in the present tense root may alternatively have j insertion (e.g. syngja).
Strong verbs
Weak verbs
1. Explain the following forms: 3rd pl. past indic. ur›u, from ver›a;
1st sg. past indic. hné, from hníga; imp. statt, from standa; pp.
ekit, from aka.
2. Why can sverja (past indic. root sór-) be said to belong to the same
strong verb type as fara (past indic. root fór-)?
3. Which pp.s of type 3 and 4 strong verbs have root vowel u rather
than o?
4. What is the past tense root of weak verbs benda, hitta, myr›a and
skipta, and why?
5. What is the past tense root of weak verbs fœ›a and ry›ja, and
why?
6. Give the 1st pl. present indic. of búa.
7. Give the 2nd sg. past indic. of strong verbs láta and slá, and the
2nd sg. present indic. of preterite present vita.
8. Give the 3rd sg. present indic. of fregna, skilja, vaxa.
9. Enumerate the different present tense roots of type 2 and type 3
strong verbs.
10. What variations does the imperative form exhibit?
11. In what way are the paradigms of frjósa and kjósa unusual?
12. Give the principal parts of koma and sofa, including all alternative
forms.
presented, and vera and sjá, since not only are these two irregular and
extremely common, but certain of their forms are easily confused.
Finally, the paradigms of one strong (fara) and one weak verb (berja)
are repeated with the -sk suffix added. Finite forms precede non-finite.
The past participle is given in the nom./acc. n. sg. form. Rather than
the abstract ‘1st sg.’ etc., pronouns are used to indicate person and
number; hann ‘he’ is used for the 3rd sg., vér for the 1st pl., flér for the
2nd pl., fleir for the 3rd pl. The imperative is always 2nd sg. (cf. 3.6.3).
Imperative skjót
Infinitive skjóta
Present participle skjótandi
Past participle skotit
Verb inflexions and their function 173
Imperative far
Infinitive fara
Present participle farandi
Past participle farit
Imperative ber
Infinitive berja
Present participle berjandi
Past participle bart/barit
174 Morphology and syntax
Imperative flakka
Infinitive flakka
Present participle flakkandi
Past participle flakkat
Imperative brenn
Infinitive brenna
Present participle brennandi
Past participle brennt
Verb inflexions and their function 175
Imperative ver
Infinitive vera
Present participle verandi
Past participle verit
Imperative sé
Infinitive sjá
Present participle sjándi
Past participle sét
176 Morphology and syntax
Preterite present verb: mega ‘be able to’ ‘be allowed to’ ‘can’
Imperative lacking
Infinitive mega
Present participle megandi
Past participle mátt/megat
Infinitive farask
Past participle farizk
Verb inflexions and their function 177
Infinitive berjask
Past participle barzk/barizk
Imperatives and present participles with the -sk suffix are uncom-
mon and in many verbs unattested. To the extent they occur, they will
be found to consist of the basic form + sk, e.g. dvelsk ‘stay!’, dveljandisk
‘staying’, from dvelja (weak type 1).
(2) Hann hljóp fyrir bor› ok svam til lands ok bjó svá um
í hvílu sinni, at flar s‡ndisk, sem ma›r lægi
‘He jumped over side and swam to land and arranged thus
around in bed REFL. POSS., that there showed-sk as man lay’
‘He jumped overboard and swam to shore and arranged his
bed in such a way that it looked as though a man lay there’
(19) Heyr flú dróttinn bœn flá, er flræll flinn bi›r flik í dag,
at augu flín sé upp lokin ok eyru flín heyrandi yfir hús
fletta dag ok nótt
180 Morphology and syntax
‘Hear you, Lord, prayer that which servant your asks you
to(-)day, that eyes your be up opened and ears your hearing
over building this day and night’
‘Hear, O Lord, the prayer which your servant asks of you to-
day, that your eyes be opened and your ears listening over this
building day and night’
(22) Vel má ek gøra flat til skaps fƒ›ur míns at brenna inni
me› honum, flví at ek hræ›umk ekki dau›a minn
‘Well can I do that for pleasure of-father my to burn inside
with him, therefore that I fear-sk not death my’
‘I can happily please my father by burning alive in the house
with him, because I do not fear death’
(24) Sér fiórr flá, at flat haf›i hann haft of nóttina fyrir skála
‘Sees fiórr then that that had he had during night-the for house’
‘Then fiórr sees that that was what he had been using during
the night as a house’
3.7 Prepositions
Occasionally of is construed with the dative case, either in sense (b) or with
the locational meaning ‘over’ ‘above’ (e.g. konungr sat of bor›i ‘the king sat
over [i.e. at] table’). The latter usage is one of shares with the prepositions um
and yfir (see below). In most functions of and um are interchangeable, and of
was more or less ousted by um, and to a lesser extent yfir, in the course of the
thirteenth century.
Like of, um may occasionally be construed with the dative, either in sense (b)
or, rarely, with the locational meaning ‘over’ ‘above’.
umhverfis ‘around’
innan ‘within’
Each of the three above prepositions can denote time as well as location (e.g.
innan lítils tíma ‘within a short time’, milli jóla ok fƒstu ‘between Christmas
and Lent’).
The above uses of at can be temporal as well as locational (e.g. lei› at kveldi
‘it passed on to evening’, at jólum gaf jarl honum gullhring ‘at Christmas the
earl gave him a gold ring’). In addition at can signify future time (e.g. at vári
‘next spring’ ‘when spring comes’).
(c) ‘from’
At + acc. in the sense ‘after’ (particularly ‘after someone’s death’) may also
be encountered (e.g. sonr á arf at taka at fƒ›ur sinn ‘a son is to take inheritance
after his father’). Historically this appears to be a different preposition from at
+ dat., probably an assimilated form of apt, related to eptir (see below).
As distinct from af, frá does not correlate with particular locational preposi-
tions, but denotes source or origin of any kind. It can have temporal as well as
locational function (e.g. frá flessum degi ‘from this day’).
gagnvart/gegnvart ‘opposite’
(b) ‘towards’
(c) ‘past’
In sense (a) and, to a certain extent, (b), ON hjá corresponds to French chez.
(b) ‘towards’
nær ‘near’
Nær can have temporal as well as locational sense (e.g. nær aptni ‘near
evening’). Since nær is in origin an adverb, it has comparative and superlative
forms (cf. 3.5.2), and occasionally these are also used with prepositional func-
tion (e.g. nær(r) honum ‘nearer him’, næst hinum fremstum ‘closest to the
foremost (people)’).
Prepositions 189
Where one entity is moving and another following or due to follow in orderly
fashion, undan corresponds to English ‘ahead of’ ‘before’ (e.g. fara undan
fleim ‘go ahead of them’).
classical Old Norse the two prepositions have become somewhat mixed
up and the one can rather often be found with the sense and/or case of
the other. In connection with the motion : location dichotomy it is
worth noting first that the movement or rest involved is often denoted
or suggested by a word other than the preposition (usually a verb or
adverb, cf., e.g., á + acc. (a) and á + dat. (a) below), and second that an
English speaker’s conception of movement and rest may not always
tally with that of speakers of other languages (cf., e.g., fleir sá bo›a
mikinn inn á fjƒr›inn ‘they saw a great breaker [i.e. breaking wave] in
the inlet [literally: (looking) into the inlet]’).
When used in a temporal sense á tends to trigger the accusative where the
noun is accompanied by the definite article (contrast dat. (b) below). Á + acc.
may indicate a point in time as well as a recurring period (e.g. á laugardaginn
næsta ‘on the next Saturday’).
Note the contrast between the accusative yfir á Nes, where the adverb yfir
indicates motion towards a place, and the dative á báti, which implies loca-
tion. The verb reri combines with both senses.
Prepositions 191
Á + dat. may indicate a point in time as well as a recurring period (e.g. á flví
sumri ‘in that summer’).
Á in this sense is typically used of body parts, but can also be found in other
contexts (e.g. allar dyrr á húsunum ‘all the doorways of the buildings’).
Eptir + acc. can also be used in the sense of ‘after someone’s death’ (e.g. flá
tók hann arf eptir fƒ›ur sinn ‘then he took inheritance after his father’).
The sense ‘following’ can extend to ‘along’ (e.g. gekk hann aptr eptir skipinu
‘he walked back along the ship’), and to ‘according to’ (e.g. gekk allt eptir
flví, sem Hallr haf›i sagt ‘everything went according to what [literally: that
which] Hallr had said’).
192 Morphology and syntax
(b) ‘ago’
Í in the above sense may be used with abstract as well as concrete nouns (e.g.
kominn í allmikla kærleika vi› ‘come into very great friendship with [i.e. be-
come very great friends with]’).
Í in this temporal sense is commonly used with the words dagr ‘day’ and nótt
‘night’ as well as the names of parts of the day and the seasons to indicate
‘time now’ or ‘time closest to the present’ (e.g. í nótt ‘tonight’, í kveld ‘this
evening’, í sumar ‘this summer’).
í + dat. ‘in’
Í in this sense may be used with abstract as well as concrete nouns (e.g. í
miklum kærleikum vi› ‘in great friendship with’).
Me› here implies that ‘he’ took his wife and children to Iceland rather than
simply going together with them (see me› + dat. (a) below). Because it carries
the notion of ‘control’ over whatever entity one is ‘with’, me› + acc. is com-
monly found with nouns denoting inanimate objects (e.g. kom Bár›r eptir
fleim me› horn fullt ‘Bár›r came after them with a full horn’, i.e. carrying a
horn full of drink).
194 Morphology and syntax
(d) ‘among’
Case usage after me› is more fluid than the above examples suggest. In par-
ticular it is not uncommon to find me› + dat. in what appears to be the ‘control’
sense (cf. me› + acc. above). As indicated in the preamble to this sub-section,
me› can sometimes take the place of vi›; thus we may attest, for example,
berjask me› + acc. for earlier berjask vi› + acc. ‘fight with’ ‘fight against’,
where the noun phrase following the preposition denotes the goal of the action.
In either of the above senses undir may be used metaphorically (e.g. gefa
undir kirkjuna ‘give to the church’, i.e. with the result that what is given comes
under the church’s control, undir fleim biskupi eru ellifu hundru› kirkna
‘under that bishop are eleven hundred churches’).
This use of vi› may be temporal as well as locational (e.g. vi› sólarsetr ‘at
sunset’, vi› fletta ‘at this (point)’).
The noun following vi› in sense (b) denotes the entity at which an action is
directed. The usage is commonly found inter alia with verbs of saying (e.g.
tala vi› konung ‘speak to the king’). Because of the directional sense, the
noun following vi› can sometimes have the force of a direct object (the ‘goal
of the action’, cf. 3.1.5, sentence 5). That is particularly the case with phrasal
verbs (those consisting of two or more words), although many of these denote
mental processes rather than actions (e.g. fara til fundar vi› ‘go to meeting
with’, where the sense is more or less equivalent to the English transitive
phrase go to meet, ver›a varr vi› ‘become aware of’, equivalent to notice,
vera hræddr vi› ‘be afraid of’, equivalent to fear).
Sometimes vi› + dat. may have the related sense ‘towards’ (e.g. horfa vi›
‘look towards’). As noted above, vi› and me› have become confused, and we
may thus find vi› + dat. in all the senses of me› + dat. Potential ambiguities
can usually be resolved by examining the sentence in which the preposition
196 Morphology and syntax
phrase stands or the wider context (e.g. slá honum ni›r vi› steininum must
mean ‘throw him down against the rock’ rather than ‘strike him down with
the rock’ because of the dative honum — cf. (3.1.5, sentence 20) that verbs of
throwing take the dative of the entity thrown).
In either of the above senses yfir may be used metaphorically (e.g. hafa vƒxt
yfir a›ra menn ‘have growth beyond [i.e. be taller than] other men’, konungr
yfir Englandi ‘king over England’).
(b) ‘without’
The only preposition to trigger all three cases is án. The meaning is
the same, irrespective of case.
án + acc./gen./dat. ‘without’
3.8 Conjunctions
Conjunctions differ from most other words in that they do not form
part of a sentence, but stand outside it. Their function, as the term
conjunction suggests, is to join constituents together, and the constitu-
ents may be anything from sentences to single sentence elements
(though even the latter can mostly be analysed as reduced sentences).
A distinction is made between coordinating and subordinating
conjunctions. The former join together constituents of the same level,
the latter constituents of different levels. Commonly, coordinating
conjunctions connect independent sentences (also known as matrix
sentences or main clauses), i.e. sentences that can stand on their own
as a complete utterance. The two most frequently occurring coordi-
nating conjunctions in English are and and but. In:
and connects the sentences Peter sat down and [Peter] poured himself
a drink (in the second sentence Peter is omitted because unless speci-
fied, the subject of poured will be understood to be the same as the
subject of sat down). Both these sentences are independent in the sense
that they require nothing further to complete them. The conjunction
but, while introducing an element of contrast not present in and, func-
tions in a similar way. In:
the sentences joined together are again both independent: (1) Anne
opened the door; (2) [Anne/she] did not go in.
Subordinating conjunctions typically function as connectors between
independent and dependent sentences, introducing the latter (also
known as embedded sentences or subordinate clauses). Dependent
sentences are those that cannot stand on their own as complete utter-
ances. Examples of subordinating conjunctions in English are because,
when, if. In:
I like the summer. The former, unlike the other sentences so far ad-
duced, is not a complete utterance. To say because it is light, which
details a reason, requires that we specify the action, event or state to
which the reason applies. Equally, the dependent when he arrives,
introduced by when, needs to be completed by an independent sentence
which details an action, event or state that stands in a time relation to
‘his’ arrival, e.g.:
In such cases that is, of course, not interchangeable with who(m), which.
We are dealing here with three fundamentally different types of de-
pendent sentence: (1) those reducible to an adverb (e.g. . . . when he
arrives = then); (2) those reducible to an adjective (e.g. . . . which you
can win = winnable); (3) those reducible to a noun phrase (e.g. . . . that
it was interesting = it, the (following) thing, etc.). All dependent sen-
tences are reducible in this way, which accounts for their dependent
status. They represent expanded versions of adverbial, adjectival or
nominal elements in independent sentences. Different though the three
main types of dependent sentence may be, it is unhelpful to divide the
words that introduce them into three separate categories since their
common function as dependent sentence introducers is thereby
obscured. In keeping with this view, all Old Norse words that join
sentences together will in the following be treated as conjunctions.
Eptir flat fór fiorfinnr jarl til Orkneyja ok sat flar um vetrinn
‘After that went fiorfinnr earl to Orkneys and sat there during
winter-the’
‘After that Earl fiorfinnr went to the Orkneys and stayed there
over the winter’
Nú mun fa›ir minn dau›r vera, ok hefir hvárki heyrt til hans
styn né hósta
‘Now will father my dead be, and has neither heard to him
groan nor cough’
‘Now my father must be dead, and neither a groan nor a cough
has been heard from him’
Although né joins together the two nouns styn and hósta, the second of these can
be seen as a reduced sentence: hefir hvárki heyrt til hans styn né [hefir heyrt til
hans] hósta.
204 Morphology and syntax
As is the case with many languages, Old Norse boasts far more subor-
dinating than coordinating conjunctions. The field is so broad, not
least because of the tendency for adverbs to metamorphose into sub-
ordinating conjunctions, that no attempt can be made here to provide
a complete list. For the student the most important thing is in any case
not the meaning of each individual conjunction. That can be looked
up in a dictionary. It is rather to grasp those features of the system that
constantly recur, in particular any which may not be immediately trans-
parent to the learner.
(a) Ok flá er flessi or›sending kom aptr til jarls, bjó hann fer›
sína
‘And then COMP this message came back to earl, prepared he
journey REFL. POSS.’
‘And when this message got back to the earl, he made ready
to leave’
Observe that the temporal contexts in which er operates are not re-
stricted to past-time reference.
Location is another type of context in which er is commonly to be
found, usually in combination with a locational adverbial. We find
flar er ‘where’ (adverb flar ‘there’), flangat er (sometimes flangat til
er) ‘to where’ (adverb flangat ‘thither’, adverb til ‘to’ (cf. 3.7.7)), fla›an
er ‘from where’ (adverb fla›an ‘thence’), and the further series hvar(gi)
er ‘wherever’ (adverb hvar ‘where’), hvert(ki) er ‘to wherever’ (adverb
hvert ‘whither’), hva›an er ‘from wherever’ (adverb hva›an ‘whence’).
Conjunctions 207
Gu› heyrir bœnir várar, hvar er vér bi›jum fyrir oss af ƒllu
hjarta
‘God hears prayers our where COMP we pray for ourselves of
all heart’
‘God hears our prayers wherever we pray from our whole heart’
Tƒl›u fleir flat órá› at leggja til bardaga vi› fiorgeir, flar er
hann haf›i li› meira
‘Said they that bad-counsel to go to battle with fiorgeirr, there
COMP he had force bigger’
‘They said it was a bad idea to go to battle with fiorgeirr
since he had the bigger force’
Muntu ok eigi vilja vita flat á flik, at flú liggir hér sem kƒttr í
hreysi, flar er ek berjumk til frelsis hvárumtveggjum
the antecedent and its accompanying pronoun are in the genitive (gov-
erned by beiddisk), but what of the correlate? The student may ponder
two possibilities: ‘which had owned Earl Einarr’ (‘which’ = subject)
or ‘which Earl Einarr had owned’ (‘which’ = object). The latter will
be preferred as by far the more likely statement, and any residual doubt
Conjunctions 209
can be resolved by the form Einarr. Einarr is nom. (acc. Einar). Since
eiga ‘[to] own’ is construed with nom. subject and acc. direct object,
the subject of the relative sentence must be Einarr, leaving the correlate
of fless flri›jungs as the object (accusative, though unmarked as such).
The idiomatic translation is thus:
In:
‘The men who were given quarter were then pulled out’
(er (or an understood correlate, cf. above) = instrumental dat. ‘with which’)
For example:
Mun ek veita flér slíkt li›, sem flú vill, at fletta fari fram
‘Will I give you such aid as you want that this goes forward’
‘I will give you as much aid as you want so that this may be
accomplished’
where at vér vindim segl várt, the object of ræ›, is represented by flat
in the independent sentence, and at flú takir vi› fiórólfi, the subject of
er, by sú (agreeing with bœn, f.).
At regularly combines with other words to form subordinating con-
junctions; these introduce various kinds of adverbial sentence. More
often than plain at the purpose conjunction is til fless at ‘in order that’
‘so that’ (til fless at can alternatively, but less commonly, have the
temporal sense ‘until’). Sentences of reason or cause may be intro-
duced by flví at, af flví at, fyrir flví at, me› flví at ‘because’ ‘since’; of
concession by fló at or flót(t) (the latter a compound of the former)
‘although’ ‘even though’; of result by svá at ‘so that’ ‘with the result
that’; and of comparison by svá . . . at ‘so . . . that’. Examples are:
Skulu [3.6.9.1 point (13)] vér frændr flínir veita flér styrk, til
fless at flú komir aldrigi sí›an í slíkt ƒngflveiti
‘Shall we kinsmen your give you support to that that you
come never subsequently into such straits’
‘We your kinsmen will give you support so that you never
again get into such straits’
214 Morphology and syntax
Kallar hann flat meirr verit hafa fyrir flví játtat, at fleir váru
flá komnir í greipr Óláfi konungi
‘Calls he that more been have for that agreed that they were
then come into clutches [belonging] to-Óláfr king’
‘He says it was agreed more readily because they had then
fallen into the clutches of King Óláfr’
Lƒg›u fleir á flótta, svá at fá ein skip váru eptir me› jarls
skipi
‘Set-off they to flight, so that few only ships were behind
with earl’s ship’
‘They took to flight, so that only a few ships were left with
the earl’s ship’
Ekki eru fleir enn svá nær oss, at eigi væri betr, at ek hef›a
sofit
‘Not are they yet so near us that not were better that I had
slept’
‘They are not yet so near us that it would not have been bet-
ter if I had slept’
Note that conjunctions that consist of more than one word can be dis-
continuous (fyrir flví . . . at; comparative svá . . . at is always so).
Conjunctions 215
Note that af flví here is the adverbial ‘therefore’, and does not belong
with the following at, which introduces the noun sentence subject of
er oss nau›syn (‘that we purify the churches of our hearts is to us a
necessity’). Observe also a further example of a correlate in a relative
sentence governed by a preposition: er hann styggvisk vi› (3.8.2.1).
hvé, hversu, hvernig ‘how’, nær, hvenær ‘when’, hví ‘why’, introduce
noun sentences. Typically such sentences occur after verbs of ‘ask-
ing’ or ‘knowing’, denoting the thing asked or known, but they may
be found in many other contexts. Since these interrogatives are among
the most common words in Old Norse and their meaning is usually
clear, the dependent sentences they introduce are unlikely to cause the
learner many difficulties. It is worth noting, however, that the pro-
nouns always appear in a case, gender and number appropriate to their
function in the dependent sentence. A selection of examples follows
to illustrate the range of Old Norse ‘indirect questions’ — as depend-
ent sentences introduced by interrogatives are often called.
Hverr is the subject of the verb ré›i and thus nominative (cf. 3.1.5, sentence 1).
The noun sentence hverr fyrir eldinum ré›i is governed by the preposition
eptir (3.7.4) in the independent sentence: if hverr fyrir eldinum ré›i were
reduced to a noun or pronoun, its case would be dative (e.g. hann lét frétta
eptir flví ‘he had people ask about that’).
Styrk, with which hvern agrees, is accusative — the object of veita (what
‘they’ (may) give). As in the first example, the noun sentence is governed by
the preposition eptir.
Conjunctions 217
Hvat, like hvern styrk in the preceding example, is the object of veita. Observe
that the noun sentence is anticipated by (and reduced to) flat in the independ-
ent sentence: ‘I want to know that — namely, what you will give us’. Such
anticipation by a demonstrative pronoun is not uncommon. fiat is acc., be-
cause it and the noun sentence it stands for are the object of vita.
Hvárum agrees with flokki, which is dat., governed by the preposition í (3.7.4).
The noun sentence introduced by í hvárum flokki is the subject of the inde-
pendent sentence — in which of the two parties the speaker finds himself is
what does not seem to him to matter.
Hva›an, nær and hvárt are interrogative adverbs and thus not inflected. The
noun sentences they introduce are the objects of veit/vita — what the speaker
does not know, what no one knows, and what the speaker wants to know.
We have already seen that the particles er and at can introduce adver-
bial sentences (3.8.2.1, 3.8.2.2). There are in addition several conjunc-
tions with more specific meaning that perform this task. Commonest
among these are the conditionals ef ‘if’, nema, útan ‘unless’, the
temporals á›r ‘before’, unz ‘until’, the temporal and circumstantial
sí›an ‘since’ ‘seeing that’ (see 3.8.2.1), and the comparatives en ‘than’,
sem ‘as’ ‘as though’. The following examples illustrate typical usage.
(a) En ef vart ver›r vi› vára fer›, flá látum vér enn hafit gæta vár
‘But if aware becomes of our movement, then let we again
sea-the guard us’
‘But if people notice our movements then we will once again
let the open sea hide us’
(a) provides a good illustration of the way in which dependent sentences are
reducible to a single word. The adverb flá ‘then’ ‘in that case’, which heads
Conjunctions 219
the independent sentence, encapsulates and repeats the adverbial sense of the
preceding conditional en ef vart ver›r vi› vára fer›. Observe further that the
conditional sentence has no subject (cf. 3.9.3).
(f) Einarr haf›i verit me› Óláfi Svía konungi sí›an Sveinn jarl
anda›isk
‘Einarr had been with Óláfr of-Swedes king since Sveinn
earl died-sk’
‘Einarr had been with Óláfr, the Swedish king, since Earl
Sveinn died’
(h) fieir létu ok eigi fleiri menn sjá á skipinu en jarli hƒf›u fylgt
‘They let also not more men see on ship-the than earl had
followed’
‘Nor did they let more men be seen on the ship than had
accompanied the earl’
(i) Hann létti eigi fyrr en hann kom á fund Magnúss konungs
‘He stopped not earlier than he came to finding of-Magnús
king’
‘He did not stop before he found King Magnús’
(j) Konungr bau› honum me› sér at vera, svá lengi sem honum
líka›i
‘King invited him with self to be as long as him pleased’
‘The king invited him to stay with him as long as he pleased’
(k) Muntu gƒrr sekr, slíkir menn sem hér eigu eptirmæli
‘Will-you made outlawed, such men as here have prosecution’
‘You will be condemned to outlawry, seeing what kind of
men have to follow up the case’
fiat segja sumir menn, at hann yr›i aldri sami ma›r ok á›r
‘That say some men that he became never same man and before’
‘Some men say that he was never the same man as (he was)
before’
and:
may either be the question ‘have you a big force?’ or the declarative
‘you have a big force’. Normally the context will make clear how such
a sentence is to be understood. Ambiguity can also be avoided by the
use of the question introducer hvárt (in origin nom./acc. n. sg. of the
interrogative pronoun hvárr ‘which of two’, cf. 3.2.5). Thus, while
may either be the question ‘is he still alive?’ or the statement ‘he is
still alive’,
However, where some other word is in first position (a) or the verb is
first (b), the order will be verb — subject — object (cf. above):
The position of the direct and indirect object in relation to one an-
other is not fixed, morphological case (mostly) indicating the function
(see 3.1.5, sentences 5, 16–19). A tendency for the indirect to precede
the direct object is however noticeable.
The subject complement (3.1.5, sentence 1) also follows the subject
in unmarked word-order, and the object complement (3.3.6, sentence 7)
the object, as in:
Points of syntax 225
Many sentences will of course contain more than subject, finite verb,
object(s) and/or complement. However, the learner is unlikely to be
much confused by the order in which such additional elements appear,
even though this can vary considerably. Three features are worth noting.
First, non-finite verb-forms may follow as well as precede objects and
complements. E.g.:
Third, provided the subject is the only noun phrase in the sentence, it
may be postponed to the end. E.g.:
the accusative case of Ásu dóttur sína should warn the student against
trying to interpret it as subject, notwithstanding it is the first noun
phrase in the sentence. The student will either know, or can ascertain
from a dictionary, that gipta is construed with a nominative subject,
accusative direct object and dative indirect object. Since Gu›rø›i
konungi is clearly dat., and Ásu dóttur sína clearly acc., hann must be
nom. (rather than acc., cf. 3.2.1) and is thus the only candidate for
subject. On the other hand, in:
both vitr ma›r and flú (ertu = ert flú, cf. 3.2.1) are nominative. There
is nevertheless no doubt that flú is subject and vitr ma›r subject com-
plement. In sentences of the X is Y type X is the topic and Y the com-
ment. Thus, you are a wise man is acceptable English since you can be
interpreted as an established discourse topic about which something
Points of syntax 227
1. What sentence positions may the finite verb occupy in Old Norse?
Give three examples.
2. In what order do the different noun phrases appear in an unmarked
Old Norse sentence? Give three examples.
3. What is meant by fronting? Give three examples.
4. Analyse the word-order of the following sentences:
(a) Fé flat allt gaf hann li›smƒnnum sínum
‘He gave all that wealth to his followers’
(b) Fornjótr hefir konungr heitit
‘There was a king called Fornjótr’
(c) Hug›u fleir, er fyrir váru, at Rƒgnvaldr jarl myndi flar fara
‘Those who were present thought that Earl Rƒgnvaldr would be on
the move there’
Such discontinuity should not on the whole cause students too much
difficulty provided they pay proper attention to case, gender and
number. The inflexions of nouns, pronouns and adjectives will nor-
mally suffice to make clear what belongs with what.
230 Morphology and syntax
correspond to passive:
Students should take careful note of these and the other types of
‘impersonal’ construction mentioned above. By one means or another
they will have to supply a subject when translating them into English.
The designation ‘impersonal’ has further been applied to Old Norse
verbs construed without a nominative, or where the nominative noun
phrase is not the first in unmarked word-order (see 3.9.1, 3.1.5,
sentence 1). This is a moot point. Where there is no nominative, there
is no person agreement in the verb — the default 3rd sg. being used
(see above); to that extent ‘impersonal’ might be deemed an appropri-
ate term. On the other hand, it has been shown that oblique (non-
nominative) noun phrases that appear first in unmarked word-order
behave like subjects in virtually every respect except the triggering of
person agreement. And such phrases may certainly denote ‘persons’.
While the question how constructions of this type are best described
is not of primary concern to the learner, it is important for him/her to
realise that where a noun phrase in a case other than the nominative is
the first in a sentence, it is not automatically to be taken as a fronted
object (see 3.9.1). Thus, in:
the accusatives hana and mik and the datives konungi and y›r are the
first noun phrases in sentences whose word-order is not obviously
marked. Even in líka›i y›r vel Finnskattrinn, where the second noun
phrase is nominative, the difficulty of showing that dat. y›r has been
fronted makes it hard to cast it in the role of object, and that in turn
raises doubts about whether Finnskattrinn can be subject. In semantic
terms, hana, mik, konungi and y›r represent ‘experiencers’ (the peo-
ple experiencing the events denoted by the verbs), a sense regularly
conveyed by the nominative in modern English and certain other
European languages (cf. I lack, I remember, I get warm, I am pleased)
— seemingly reflecting a common tendency to make the experiencer
subject rather than the thing experienced. Certainly, natural English
translations of Old Norse sentences like the above will tend to bring
out the subjecthood of the first noun phrase.
Also regularly construed without nominatives are the passives of
verbs whose direct object is in the genitive or dative, e.g.:
In the active, leita ‘seek’ ‘ask’ has a nominative subject and genitive
object, bjarga ‘save’ a nominative subject and dative object. When
passivised such verbs lose their nominative subject in the normal way
(3.6.4), but the object does not become the new nominative subject. It
remains in its original case. However, since in unmarked word-order
(cf. 3.1.5, sentence 1) it precedes the verb phrase in the passive sentence,
there is some justification for treating it as subject. It certainly becomes
the theme of the sentence — ‘what it is about’ (cf. 3.1.5, sentence 1).
Only partially analogous are passives of verbs construed in the active
with a nominative subject, a dative indirect object and a further argu-
ment in the genitive or dative. To active:
Points of syntax 235
correspond passive:
In English we may say: I saw her open it, he asked the boys to sing.
What follows saw and asked is sometimes described as a non-finite
clause object: we have a clause or sentence which is the equivalent of
an object (cf. I saw the letter, he asked a favour), and it contains an
Points of syntax 237
infinitive (open, sing) but no finite verb. This analysis, however, leaves
out of account the fact that in a sense her is both the object of saw and
the subject of open, and the boys both the object of asked and subject
of sing (cf. (I saw that) she opened it, (he asked the boys that) they
should sing). The term mostly used to describe the Old Norse counter-
parts of such English constructions is ‘accusative and infinitive’. While
hardly achieving descriptive adequacy, this designation has the merit
of emphasising accusative case, which marks direct object status, and
suggesting a connection between the accusative and the following in-
finitive. Above all, it is a more precise term than non-finite clause
object, which can cover a variety of constructions.
Old Norse accusatives and infinitives occur regularly after verbs of
saying, thinking, and experiencing. E.g.:
the -sk can be interpreted literally as ‘herself’. Observe that the sub-
ject complement dóttir is nominative. This is the rule where the accu-
sative of an acc. + inf. construction is to be found in the -sk suffix.
238 Morphology and syntax
In (a) nom. fleir is the subject of flóttu as can be seen from the 3rd pl.
verb-form. In (b), on the other hand, where flótti is 3rd sg., fleir can
only be the subject of inf. sœkja. When the subject of the infinitive is
3rd sg., as it often is, the two constructions are difficult to distinguish. In:
It is difficult to get the literal sense of mundu and skyldu across since
English ‘would’ and ‘should’ are finite forms. Semi-literal renderings
may be helpful here, using the infinitive marker to to direct attention
to the past infinitive.
3.9.5 Omissions
Certain elements are regularly omitted from Old Norse sentences. Some
can be readily understood from the context and will cause the learner
no difficulty. A subject that is already established, for example, is usu-
ally omitted in Old Norse just as in English. Thus, in:
the subject of ba› is not expressed — any more than in the English
renderings — because it refers to the same person as the subject of the
previous sentence, Karl.
Points of syntax 241
3.9.5.1 Objects
Here ‘it’, referring in (a) to the bag of money, in (b) to the sword, is
lacking in Old Norse. This is because there is identity of reference
with a preceding noun, fésjó›inn in (a), sver›s in (b). Observe that
object omission is not dependent on case equivalence. In (a) the missing
noun phrase would have had accusative case, just as fésjó›inn, but in
(b) it would have been dative, while sver›s, the noun establishing the
reference in (b), is genitive, governed by the preposition til (3.7.2).
Indirect objects, too, may be omitted, as in:
Note that the idiomatic English rendering obscures the omission; in-
sertion of ‘him’ between ‘taught’ and ‘the’ would give a different sense
— that it was the king himself who taught Hákon the true faith.
3.9.5.2 vera
The verb vera is often omitted, especially the infinitive (a) in connec-
tion with auxiliary verbs and (b) in accusative and infinitive construc-
tions. The student should pay particular attention to this phenomenon
since it can often cause misunderstanding.
242 Morphology and syntax
In the first example the copula (vera, the verb ‘be’) is the missing link
needed to connect subject flú and the subject complement vel kominn
(cf. flú ert vel kominn ‘you are welcome’). In the second vera is required
to complete the passive construction leitat skyldi vera ‘should be
sought’. The third and fourth examples illustrate accusative and infinitive
constructions from which the infinitive is omitted. In the third the
copula is what is wanted to connect flat and its complement ósannligt
(cf. flat er ósannligt — subject + copula + subject complement), so the
construction is to be understood as fiorfinnr kva› flat ósannligt vera,
at . . . The fourth example too requires vera to be understood since
várkunn can only be the object of tƒl›u in an accusative and infinitive
construction (i.e. tƒl›u sumir várkunn vera at . . . is perfectly accept-
able, but *tƒl›u sumir várkunn without the ellipsis of vera is not).
Because all four contexts so clearly demand vera, it is readily under-
stood or supplied by the reader familiar with Old Norse. The beginner
will have to proceed more slowly and analytically: faced by a sentence
that seems to lack an infinitive, and in doubt about the meaning, s/he
should always try supplying vera. In most cases this will provide the
solution.
Points of syntax 243
Finite forms of vera are also sometimes omitted. As with the above,
the prerequisite seems to be that the verb should be recoverable from
the context. Consider:
The finite verb of the first sentence is hƒf›u, but that will not fit the
context of the second. What we have in sárir allir a›rir is a fronted
subject complement (sárir) followed by the subject (allir a›rir), and
the copula is needed to connect them. The second sentence is thus to
be understood: en sárir váru allir a›rir.
Important aspects of nominal syntax not dealt with elsewhere are (1)
certain idiomatic uses of personal pronouns and possessive adjectives;
(2) what are often loosely termed ‘the genitive and dative of respect’.
Here, as will be seen from the idiomatic translations, the pronouns are
only partly in apposition to the personal names since they also contain
a reference to one or more other people known from the context. The
dual pronouns denote one additional person, the 1st and 2nd plural more
than one. Thus vér Arnvi›r would mean ‘Arnvi›r and we (others)’, flit
Arnvi›r ‘Arnvi›r and you [sg.]’, flér Arnvi›r ‘Arnvi›r and you (others)’.
Since there is no dual 3rd person pronoun, fleir Rƒgnvaldr can mean
‘Rƒgnvaldr and he’ as well as ‘Rƒgnvaldr and the others’, depending
on the context. Where men and women or a man and a woman are
involved, the 3rd person neuter plural is used (cf. 3.2.1):
From the context of this particular example we know that only the
queen and a single male are involved; in another context flau dróttning
could mean ‘the queen and the others (including at least one male)’.
This usage is not confined to personal pronouns, but can also be
found with possessive adjectives. E.g.:
Here the dual 2nd person possessive ykkra carries the same ‘inclu-
sive’ sense as the personal pronouns in the previous examples. There
is however a significant syntactic difference between ykkra Óláfs digra
and, say, vit Arnvi›r. The pronoun vit stands in the same case as Arnvi›r
(nom.), whereas ykkra takes its case (and gender and number) from
deilu (acc. f. sg.) while Óláfs digra is in the genitive. The difference is
occasioned by the fact that vit and Arnvi›r form a joint subject, a pair-
ing of two noun phrases, whereas ykkra and Óláfs, though both modify
deilu, represent different word classes: adjective and noun. The pos-
sessive signals its modifier role by case, gender and number agree-
ment, but the noun cannot — instead it goes into the genitive (the
‘possessive’ case, see 3.1.5, sentence 13). The close relationship be-
tween possessive adjectives and genitives is shown by the 3rd person
pronouns (non-reflexive), whose genitive forms, hans, hennar, fless,
fleira, double up as possessives (3.3.8.5 point (6)).
It should be observed that the juxtaposition of possessive adjectives
and genitive noun phrases is also common in more unambiguous cases
of apposition. E.g.:
Instead of nom. m. sg. hverr ‘each’ + gen. y›var ‘of you’ and nom. m.
sg. sá ‘that one’ + gen. okkar ‘of us two’, we find nom. m. sg. hverr
y›varr ‘each your’ and nom. m. sg. sá okkarr ‘that one our-two’.
Students should take careful note of this construction since experience
has shown it can cause much confusion.
Contemptuous reference is a further case in which a possessive adjec-
tive is used where on the basis of English one might expect a personal
pronoun. This can occur in both direct and indirect speech. Thus we
find not only fóli flinn ‘fool your [i.e. you fool!]’, but also:
The genitive and dative can be used in Old Norse to specify the appli-
cability of the verb phrase. The basic sense of such constructions is
‘with respect to’ ‘in respect of’ ‘as regards’, but idiomatic English
will usually require a different translation. E.g. (with the genitive or
dative phrases in bold):
A few remarks on verbal syntax need to be added to the basics set out
at various points in section 3.6. These concern four areas: (1) the per-
fect and past perfect (3.6.2, 3.6.8, sentence 3); (2) the passive (3.6.4);
(3) the ‘dative absolute’; (4) present participles expressing potential-
ity or obligation.
Svá mun Hallger›i s‡nask, sem hann hafi eigi sjálfdau›r or›it
‘So will to-Hallger›r seem, as he has not self-dead become’
‘It will seem to Hallger›r as though he has not died a natural
death’
One of the chief reasons for assuming this development is that in early
Old Norse texts the past participle quite often agrees with an accusa-
tive object (never a genitive or dative, since hafa governed the accu-
sative only). It seems, however, that at this relatively late stage in the
history of the Old Norse perfect, participle-object agreement had ceased
to carry the original ‘I have them bought’ meaning. Agreement and
non-agreement give the appearance of being interchangeable — indeed,
sometimes we find an inflected and an uninflected participle depend-
ent on the same auxiliary, e.g. (with the participles in bold):
In the first example frelsta agrees with frændr sína (acc. m. pl.), in the
second and third examples the nom./acc. n. sg. form of the participle
is used, once where the object is acc. m. sg. (steininn), once where it is
dat. reflexive.
The Old Norse periphrastic passive formed with vera may be dynamic
or static, just as its English counterpart with be. Dynamic passives
denote an action or event, static passives the state after an action or
event. Two typical examples illustrating the difference are:
In the first example the talk is of sons who may be born in the future.
The interpretation of the second example is less certain: it need be no
more than a dynamic passive, but it could carry the additional sense
that an uprising against the king was not possible on that occasion
Points of syntax 253
(because of his superior force). More firmly endowed with the notion
of (im)possibility is:
Such constructions usually carry the implication that the action was a
chance one, a sense of ver›a being ‘[to] happen’ (cf. slíkt ver›r opt
ungum mƒnnum ‘such things often happen to young men’). A more
precise idiomatic rendering of the above would therefore be: ‘He
chanced to look up at the hillside’.
Vér skulum hér koma svá margir flingmenn, sem nú eru til
nefndir, at uppverandi sólu
‘We shall here come as many assembly-members as now are
to appointed, with up-being sun’
‘We are to come here, as many assembly members as are
now appointed for the purpose, when the sun is up’
The clarity of the context will determine the degree of precision with
which the participle can be translated into English.
Outside this construction, the Old Norse present participle tends to
correspond to the English -ing form of the verb and will give the learner
little trouble: e.g. hlæjandi ‘laughing’, skínandi ‘shining’, sofandi
‘sleeping’.
1. When is hafa and when vera used to form perfect and past perfect
constructions?
2. To what in Old Norse does the term ‘supine’ refer?
3. What is the difference between an inflected past participle and an
inflected supine in Old Norse?
4. Give an example of a dynamic and a static passive in Old Norse
and explain the difference.
5. What characterises ver›a-passives?
6. What is the Old Norse ‘dative absolute’? Give examples of the
construction.
7. Explain the meaning of the present participle in: flat flótti fló
ógeranda, at konungr vissi eigi fletta.
256 Morphology and syntax
Complements of vera ‘[to] be’ and ver›a ‘[to] become’ are some-
times adverbs in Old Norse. In the case of the pair vel ‘well’ and illa
‘badly’, English tends to use adjectives in corresponding phrases. E.g.:
fiat er vel
‘It is good’ ‘It is right’
Although the verb in Old Norse normally agrees in number with the
(nominative) subject of the sentence (3.1.1, 3.2, 3.6.1), there are ex-
ceptions to the rule. Where the verb precedes one or more of a sequence
of conjoined subjects, it will often appear in the same number as the
subject which is closest. E.g. (with the relevant agreement in bold):
Observe that in the first sentence the past participle komnir agrees
with the plural subject allir stafnbúarnir (or, equally possible, both
subjects together) rather than the singular subject Ulfr and the singular
verb. (On the use of flau in the last sentence, see 3.9.6.1.)
Even where it precedes a lone plural subject, a verb may appear in
the singular if several words intervene. E.g. (with the singular verb in
bold):
As pointed out in 3.6.5.3, the -sk form may have reflexive and recip-
rocal function. Often this is combined with use of a preposition, which,
in an abstract sense, governs the reflexive or reciprocal to which the
-sk form gives expression. Since, however, there is no overt prepositional
complement in such constructions, the preposition has the appearance
of an adverb (cf. 3.7.7). E.g. (with the -sk form and preposition given
in bold):
Kormakr litask um
‘Kormakr looks-sk around’
‘Kormakr looks around him’
3.9.9 Adverbial ok
The second example is from a law text, where this use of ok for flá is
very common.
262 Morphology and syntax
The student may legitimately wonder why some verbs in Old Norse
are construed without a nominative, and thus, apparently (cf. 3.9.3),
without a subject. It was explained in 3.9.3 that sometimes this is
because the focus is on the object and the subject is of no interest in
the context. In, e.g.,
the writer draws attention solely to the work and its commencement.
Who caused it to commence is of no relevance, and indeed the
individual(s) concerned would probably be hard to identify. There is
a similarity here with some passive constructions in English. In, e.g.,
the point of interest is the time at which the building work took place,
not who carried it out, which, as in the ON example, may not be
(generally) known.
What is missing in both the ON and the English sentence is of course
the agent — which is nevertheless there in the background, understood
although unspecified. But it has been argued by some that an agent
has also been omitted from those types of ON ‘impersonal’ (i.e. non-
nominative) construction in which an animate instigator cannot be
conceived (e.g. daga ‘dawn’, skorta ‘lack’, dreyma ‘dream’; cf. further
3.9.3, pp. 230, 233–4). The verbs concerned tend to denote natural
events, the passing of time, (chance) occurrences, want, feelings, im-
pressions, etc. What is suggested is that at the time such constructions
arose there was a belief in a mythological agency or agencies which
controlled the events by which people were affected. Thus in, e.g.,
head word The central word in a phrase. E.g. in ma›r gamall ‘an old
man’, the noun ma›r is the head word, on which the adjective gamall is
dependent; we find ma›r gekk út ‘a man went out’, but not *gamall
gekk út ‘old went out’.
infinitive clause A clause (sentence) with one or more infinitives but no
finite verb — one of several types of non-finite clause. E.g. in dvel flú
eigi at snúask til dróttins ‘do not wait to turn to the Lord’, at snúask til
dróttins is an infinitive clause whose covert (understood) subject is the
flú of the finite dvel flú eigi (see covert).
infix An affix (a word element that can only be used when joined to another
form) added within a word (see p. 65).
inflexion A change to any part of a word (root, affix, ending) signalling
grammatical relations (case, gender, number, tense, mood, etc.), e.g.
hestr ‘horse’ (nom. sg.), hests (gen. sg.); hƒr› ‘hard’ (nom. f. sg., nom./
acc. n. pl.), har›ir (nom. m. pl.); bít ‘bite’ (1st sg. pres. indic.), beit (1st/
3rd sg. past indic.).
intransitive A verb which cannot take a direct object, e.g. liggja ‘lie’ ‘be
situated’.
modifier (modify) A word that is dependent on another word or phrase
and qualifies its meaning. In, e.g., sá inn gamli ma›r ‘the old man’, the
words sá inn gamli are all dependent on the head word ma›r: they qualify
the meaning of ma›r, introducing the attribute of age and making the
phrase definite. In draumr Hálfdanar, the genitive Hálfdanar is de-
pendent on draumr and qualifies its meaning, indicating whose the dream
was (see head word).
overt Expressed, observable in the structure (see covert).
past perfect A verb construction found in Germanic (and some other)
languages consisting of the past tense of have (hafa etc.) and a supine,
usually expressing a time prior to some past point of time. In, e.g., er
fleir hƒf›u upp borit fƒngin, fóru fleir á land ‘after they had carried up
the baggage, they went ashore’, the carrying precedes the going ashore,
which is itself in the past.
perfect As past perfect, but consisting of the present tense of have (hafa
etc.) and a supine, and commonly expressing time viewed in relation to
the present. In, e.g., vér hƒfum fengit mikinn ska›a á mƒnnum várum
‘we have suffered great losses to our men’, the losses are presented as
relevant to the situation in which the words are spoken.
phrase A sentence element consisting of one or more words, but usually
reducible to a single word. E.g. skrín ins helga Magnúss jarls ‘the shrine
of St Magnús the earl’ is a noun phrase, reducible to flat ‘it’; í flann
268 A New Introduction to Old Norse
suffix An affix (see infix) coming after the form to which it is joined.
Examples of ON suffixes are the weak past tense markers -›, -d, -t (cf.
kasta-›-i ‘threw’, with root kasta-, past tense suffix -› and 3rd sg. ending
-i), and the definite article -inn (cf. hest-r-inn ‘the horse’, with root
hest-, nom. sg. ending -r and suffixed nom. m. sg. def. art. -inn).
transitive Verb which can take a direct object, e.g. drepa ‘kill’, hefna
‘avenge’.
Bibliography
Cleasby, Richard and Gudbrand Vigfusson 1957. An Icelandic–English
Dictionary (2nd ed.). Oxford: Clarendon Press.
Faarlund, Jan Terje 2004. The Syntax of Old Norse. Oxford: Oxford
University Press.
Gordon, E. V. 1957. An Introduction to Old Norse (2nd ed.). Oxford:
Clarendon Press.
Haugen, Einar 1972. First Grammatical Treatise (2nd ed.). London:
Longman.
Haugen, Odd Einar 2001 (and later printings). Grunnbok i norrønt
språk (4th ed.). Oslo: Ad Notam Gyldendal.
Heusler, Andreas 1932 (and later printings). Altisländisches Elementar-
buch (3rd ed.). Heidelberg: Carl Winter.
Hreinn Benediktsson 1972. The First Grammatical Treatise. Reykjavík:
University of Iceland.
Höskuldur Thráinsson 1994. ‘Icelandic’. In: (E. König and J. van der
Auwera eds) The Germanic Languages. London: Routledge, 142–89.
Iversen, Ragnvald 1973 (and later printings). Norrøn grammatikk (7th
ed.). Oslo: Aschehoug.
Noreen, Adolf 1923. Altisländische und altnorwegische Grammatik
(4th ed.). Halle: Niemeyer.
Nygaard, M. 1905. Norrøn syntax. Kristiania: Aschehoug.
Spurkland, Terje 1989. Innføring i norrønt språk. Oslo: Universitets-
forlaget.
Stefán Einarsson 1945 (and later printings). Icelandic. Baltimore: John
Hopkins.
Thomson, Colin D. 1987. Icelandic Inflections. Hamburg: Helmut
Buske.
Zoëga, Geir T. 1910. A Concise Dictionary of Old Icelandic. Oxford:
Clarendon Press.