An Introduction to Old English Edinburgh University Press

tary distribution with the voiceless ones. That is to say, when a fricative
8
AN INTRODUCTION TO OLD ENGLISH
02 pages 001-166 29/1/03 16:09 Page 8

---

phoneme occurred at the beginning or end of a word, then it was
produced as voiceless, but in the middle of a word it was produced as
voiced. Thus the word full ‘full’ would have been phonetically [full],
and the word drı¯fan ‘drive’ would have been [dri

van]. But phonemically
both fricatives would have been /f/, i.e. /full/, /dri

fan/. The develop-
ment of a contrast between voiceless and voiced fricatives, as in standard
PDE ferry vs. very, is a feature of the Middle English period. Another
feature about the fricatives is more obvious, namely the presence of /x/,
which does not occur in PDE. This voiceless velar fricative is compar-
able to the same sound in German and Dutch hoch, hoog ‘high’ so we find
OE he¯ah. If we stay with the word he¯ah, it is worth noting that the initial
consonant, although originally [x], had changed into the glottal fricative
[h] by the OE period, thus already having the pronunciation it has in
PDE. However, phonemically it remained an allophone of the phoneme
/x/, and as we shall see below, it contrasts with the initial sound of a
word such as guma ‘man’.
As with the other fricatives, the velar is voiced medially, but excep-
tionally this sound, [
γ
], appears to have been a separate phoneme /
γ
/.
It occurs initially, as in guma ‘man’, and medially, as in dagas ‘days’. But it
does not appear finally, where the sound is voiceless, hence /x/, as in sorh
‘sorrow’. This voiced fricative is difficult for PDE native speakers to
produce, since it is foreign to the present-day sound system. Since it is
known that by the very end of the period the initial sound was develop-
ing to /g/, it makes sense to substitute that phoneme when reading.
Similarly, the medial sound was to develop later into a variety of other
sounds, and it may ease your introduction into the OE sound system if
you use /w/, especially when the etymology suggests that that is the
later state of affairs, as in boga ‘bow’.
There were two sibilant phonemes, /s/ and /
ʃ
/, but only the former
had a voiced allophone medially. Otherwise they behave in a fashion
parallel to the fricatives. I shall discuss the behaviour of /
ʃ
/ further
below. In addition to these sibilants, OE also had two affricates, namely
/
tʃ
/, as in c
.
yric
.
e ‘church’, and /
d
/, as in ec
.
g
. ‘edge’, see §1.4 for the
spelling of the affricates.
Unlike the situation in PDE, there were only two nasal phonemes in
OE, namely /m/ and /n/. The difference arises because in OE when the
phonetic sound [
ŋ
] occurs, it is always followed by either [k], as in 
†
anc
‘thank’, or [g], as in sing ‘sing’. Therefore it remains an allophone of /n/.
In standard PDE, on the other hand, final [g] has been lost, so that /
ŋ
/ is
phonemic. It is worth noting that in the English Midlands the situation
is close to the OE one, for there the final [g] has remained.
As in PDE there were two phonemic liquids in OE, namely /l/ and
ORIGINS AND SOURCES
9
02 pages 001-166 29/1/03 16:09 Page 9

---

/r/. The former was similar to that in PDE, and probably had two allo-
phones, ‘clear’ [l] initially and ‘dark’ [

] elsewhere, as in lytel ‘little’,
where in both OE and PDE the first l is clear and the second l is dark.
The one thing about /r/ of which we can be certain is that its pronun-
ciation was quite different from that of PDE /r/. It is probably imposs-
ible, at this distance, even to attempt accuracy. Perhaps a sound in the
range between an alveolar trill and a flap would be most appropriate.
Finally, and before final consonants, it may well have had a retroflex or
velarised component. Whatever the case, it must be observed that a post-
vocalic /r/ is always pronounced, in contrast to the situation in PDE.
Initial and final examples are rı¯dan ‘ride’ and heard ‘hard’.
There are two further consonants to mention, namely the approxi-
mants /j/ and /w/. Neither is particularly difficult and they are both
directly reflected in the corresponding PDE forms. Phonologically they
are the consonantal counterparts of the high vowels /i/ and /u/. The
real problems with both of them, and especially with /j/, lie in the
complexity of the OE spelling system, but see §1.4 for some help in this
area.
There are two areas where OE had distinctive characteristics which
are no longer present in PDE. Firstly, we find initial clusters consisting
of /x/ + liquid, nasal or approximant, i.e. /xl-, xr-, xn-, xw-/, as in hlu¯d
‘loud’, hring ‘ring’, hnæg