An Introduction to Old English Edinburgh University Press

Infinitive
Past Sing.
Past Pl.
Past Part.
rı¯dan
ra¯d
ridon
g
.
eriden
The most important warning to issue here is that the references to
‘past singular’ and ‘past plural’ are misleading. ‘Past singular’ refers only
to special forms for the 1st and 3rd singular of the indicative, i.e. the
equivalents of I rode and he/she/it rode. Every other past tense form
follows the shape of what I have called the ‘past plural’.
A comparison of the full paradigm of a verb such as drı¯fan with,
say, de¯man, shows a number of distinctions between weak and strong
verbs:
Present
Past
Indicative
1 Sing.
drı¯fe
dra¯f
2 Sing.
drı¯fst
drife
3 Sing.
drı¯f
e
dra¯f
Plural
drı¯fa
e
drifon
Subjunctive
Sing.
drı¯fe
drife
Plural
drı¯fen
drifen
56
AN INTRODUCTION TO OLD ENGLISH
02 pages 001-166 29/1/03 16:09 Page 56

---

Imperative
2nd Sing.
drı¯f
_____
2nd Plural
drı¯fa
e
_____
Participle
drı¯fende
g
.
edrifen
Within the present tense the differences which appear in the 2nd sing.
and 3rd sing. are more apparent than real. They arise out of the process
of syncope which I discussed when dealing with de¯man in §4.2. This
should not cause you any difficulties in the paradigm above. However,
in many other verbs, such as rı¯dan, the process of syncope would have
produced rı¯dst, rı¯d
´
. This quite often results in awkward sequences of
consonants and therefore various types of assimilation, that is to say,
the process of making different consonants more or completely similar,
occurs. Assimilation changes the sequence to rı¯tst, rı¯tt. Similarly from
‘expected’ bı¯t
´
‘it bites’ we get bı¯tt.
The real differences occur in the past tense and past participle forms.
I have already mentioned that the 1st and 3rd singular indicative forms
are quite different from any of the others. Note now that this extends to
the fact that there is no inflectional suffix at all in these two forms, which
thus contrast with the weak verbs. Of course the major contrast between
the past tense forms of weak and strong verbs is represented by the fact
that in all weak verbs there is a suffix intervening between the stem and
the inflectional ending. This suffixal ending always contains a /d/ or, less
often, a /t/. As a result it is customary to call this a dental suffix. This is
a primary feature of weak verbs which is entirely absent from strong
verbs. For strong verbs the inflectional endings are added directly to the
stem, and this accounts, for example, for the past plural form drifon,
compare de¯mdon, where the dental suffix -d- is apparent.
Another quite small difference occurs in the 2nd sing., which does not
have the -st inflection found elsewhere. Of much more interest, however,
is the past participle. In all strong verbs the inflection of this participle is
always -en. This is often, but not always, paralleled in the present-day
language. Thus, for example, the past participle of sing is today sung,
but in Old English it was g
.
esungen. Any of you who know either Dutch or
German will immediately realise that in this respect Old English, Dutch
and German are entirely parallel.
All strong verbs inflect in much the same way, regardless of which class
they belong to, but below I shall mention some of the more important
variations which are found. But the fact that, unlike the weak verbs,
strong verbs cohere as a group in terms of inflection means that we can
proceed for the moment on a different track. From everything I have said
above it should be clear that the differences between different classes of
STRONG VERBS
57
02 pages 001-166 29/1/03 16:09 Page 57

---

strong verbs are a matter of different ablaut patterns. Let us, therefore,
look at the pattern found in each class:
I
drı¯fan
dra¯f
drifon
-drifen ‘drive’
II
sme¯ocan
sme¯ac
smucon
-smocen ‘smoke’
III
singan
sanc
suncon
-suncen ‘sink’
IV
beran
bær
bæ¯ron
-boren ‘bear’
V
sprecan
spræc
spræ¯con
-sprecen ‘speak’
As it stands, the display above is unhelpful. It looks like no more than
a more complex version of what we find today. That may well be true,
but it is possible to make more sense of the situation. For although by the
Old English period the original ablaut patterns had been considerably
obscured by the passage of time (and sound change), it is possible to
reconstruct what the situation might have been like in Germanic. And in
so doing some remarkable features emerge.
Recall now that the defining feature of Ablaut is the stem, or,
more accurately, root vowel. If we take a class I verb and analyse it as
containing an onset, a nucleus and a coda, that is to say, a set of initial
consonants, a vocalic nucleus and a set of final consonants, then we can
define, say, drı¯fan, as: /dr/ + /
i
/ + /f/ (ignoring the inflection as
irrelevant). It will be obvious that the crucial element is the nucleus.
We still have some work to do, and the essential business consists of a
two-stage process. Firstly, instead of saying that there is a long vowel in
the nucleus, let us say that there are two short vowels, i.e. /i + i/. This is
quite a common analysis of long vowels in phonological theory, so it is
quite acceptable here too. The second stage will appear more mysteri-
ous, but what I propose to do is to label the two vowels differently. I shall
call the first vowel the Ablaut vowel, or A (for Ablaut), and the second
vowel a contextual element, or X. That may seem mysterious, but it
should be recalled that we have already seen that Ablaut is the defining
feature of strong verbs, and thus what I am doing is claiming that Ablaut
is not just something in the ether, but rather has overt expression.
Having done this analysis, then we can say that drı¯fan actually has an
internal structure of Onset + Nucleus + Coda and that the Nucleus
consists of Ablaut vowel + some defined context, as I shall show very
shortly. The structure of the onset is generally irrelevant, but we do need
to show any coda structure. It is true that in some cases, including class
I, it would not seem necessary to invoke any reference to the coda. That
is not wholly true even in that class, although the argument would
require me to get involved in matters to do with the earliest stage of
Indo-European. However, it will shortly become clear that elements of
58
AN INTRODUCTION TO OLD ENGLISH
02 pages 001-166 29/1/03 16:09 Page 58

---

the coda can indeed be crucial, so it seems most helpful to indroduce the
concept even here.
Let me show how the Ablaut system works in practice:
drı¯f-
dra¯f-
drif-
drif-
-AXC-
-AXC-
-AXC-
-AXC-
-eiC-
-aiC-
-ØiC-
-ØiC-
It can be seen that the contextual element is /i/ and that it remains
constant. It is this contextual element which is the defining feature of the
different verb classes. On the other hand, the Ablaut element varies,
appearing as either /e/ or /a/ or Ø (zero). The reason for this is that
Ablaut is subject to two types of gradation (hence the English term vowel
gradation). There is qualitative gradation, which occurs when one
vowel replaces another. The variation always involves variation between
/e/ and /a/, so that is the explanation for the alternation between -ei-
and -ai-. There is also quantitative gradation; in the examples above
this is realised as variation between a normal length vowel and either
Ø or a reduced vowel similar to the initial of present-day about. This
reduced vowel is often called schwa and it has some variation, but I shall
use /
ə
/ when needed. So, still schematically, we can describe the ablaut
pattern above as: front ~ back ~ reduced ~ reduced. The other variations
from Old English which can be seen above, namely the monophthong-
isation of /ei/ to /
i
/ and of /ai/ to /
a
/ are quite separate developments
in English, nothing to do with Ablaut, but belonging to the pre-history
of Old English and rather outside this book.
All this may seem rather random, perhaps to the extent of being quite
unhelpful. But compare with the Ablaut series above the corresponding
series for class II verbs:
sme¯oc-
sme¯ac
smuc-
smoc-
-AXC-
-AXC-
-AXC-
-AXC-
-euC-
-auC-
-ØuC-
-ØuC-
The only significant difference between the two series is that the con-
textual element has changed from /i/ to /u/. Of course, as before, the
situation is obscured by other Old English developments. But, rather
than concentrate on those, the general principle, that the main shift
between different classes is caused by the contextual element which
immediately follows the ablauting vowel, should be recognised as the
critical feature.
The distinguishing feature of class III verbs is that the context is
either a liquid (/l, r/) or a nasal (/m, n/), which are classed together as
sonorants (S). Thus we find:
STRONG VERBS
59
02 pages 001-166 29/1/03 16:09 Page 59

---

sing-
sang
sung-
sung-
-AXC-
-AXC-
-AXC-
-AXC-
-eSC-
-aSC-
-
ə
SC-
-
ə
SC-
Note that I have used schwa rather than zero, for here the ablaut vowel
surfaces as /u/, rather than disappearing.
So far I have suggested that there are two ablaut grades: normal and
reduced. However, it is also possible to find a lengthened grade, where
the normal vowel becomes long. This happens in classes IV and V, where
the past plural has lengthened grade (and also, qualitatively, a-grade).
Thus in class IV we find the following; note that I have specified the
context as a sonorant:
ber-
bær
bæ¯r-
bor-
-AS-
-AS-
-AAS-
-AS-
-eS-
-aS-
-eeS-
-
ə
S-
Here and in class V there are, as elsewhere, some particular Old English
developments. Otherwise class V is parallel to class IV except that the
final element is an obstruent (O), i.e. a ‘true’ consonant, rather than a
sonorant, and this affects the final shape of the schwa vowel:
sprec-
spræc
spræ¯c-
sprec-
-AO-
-AO-
-AAO-
-AO-
-eO-
-aO-
-eeO-
-
ə
O-
It would be possible to extend this analysis of Ablaut to class VI, but
this would involve an extremely complex and not necessarily rewarding
discussion. Instead, therefore, I simply present a typical example:
VI
faran
fo¯r
fo¯ron
-faren ‘go’
So far I have ignored class VII verbs, although they are often of quite
high frequency. My reason for this is that they do not fit in to the ablaut
series proper and are of varied origin. Nor is it possible, as it is with other
strong verbs, to determine the class by reference to their infinitive. The
best means of identifying these verbs is by their past tense forms, where
they all have stem vowel e¯o or, less often, e¯. Typical examples of each
time are shown below:
VII (a) feallan
fe¯oll
fe¯ollon
-feallan ‘go’
VII
(b) ha¯tan
he¯t
he¯ton
-ha¯ten ‘call’
The overwhelming majority of these verbs also have the same stem
vowel in the past participle as in the infinitive.
60
AN INTRODUCTION TO OLD ENGLISH
02 pages 001-166 29/1/03 16:09 Page 60

---

5.4 Variation in strong verbs
As I have hinted at, there are a number of areas where there is variation
both within and across strong verb classes. This should not be surprising,
for these are, historically, of longer lineage than the weak verbs, and
therefore there was greater time in which variation could arise. It is
impossible here to account for all this variation, and I shall restrict
myself to the most important of the changes.
An obvious place to start is with i-mutation. In strong verbs
i-mutation is evident in the 2nd and 3rd person singular present forms.
Of course not every strong verb is affected: for example most class I are
unaffected, since their stem vowel is ı¯. It is always found with classes II,
VI and VII, and a technically different, although in practice identical,
shift of e to i in classes IV and V has exactly the same effect, so that we
find beran ‘bear’ but birst ‘thou bearest’. Some other sound changes in the
development of the language can also cause some minor variations in the
stem vowel, so that, for example, we find class III helpan ‘help’ with stem
vowel e which then can be subject to i-mutation to give hilpst in a fairly
transparent way. I am not going to discuss most of these changes, which
would merely muddy the waters unnecessarily.
There is a small group of verbs in classes V and VI which form their
present tense in the same way as weak class 1 verbs, e.g. sittan ‘sit’ V,