The history of Old English and its development
The history of Old English and its development.
In 409 AD the last Roman legion left British shores,
and in fifty years the Islands became a victim of invaders. Germanic tribes
from Southern Scandinavia and Northern Germany, pushed from their densely populated
homelands, looked for a new land to settle. At that time the British Isles were
inhabited by the Celts and remaining Roman colonists, who failed to organize
any resistance against Germanic intruders, and so had to let them settle here.
This is how the Old English language was born.
Celtic tribes crossed the Channel and starting to
settle in Britain already in the 7th century BC. The very word
"Britain" seems to be the name given by the pre-Celtic inhabitants of
the island, accepted by first Indo-Europeans. The Celts quickly spread over the
island, and only in the north still existed non-Indo-European peoples which are
sometimes called "Picts" (the name given by Romans). Picts lived in
Scotland and on Shetland Islands and represented the most ancient population of
the Isles, the origin of which is unknown. Picts do not seem to leave any
features of their language to Indo-European population of Britain - the famous
Irish and Welsh initial mutations of consonants can be the only sign of the
substratum left by unknown nations of Britain. At the time the Celts reached
Britain they spoke the common language, close to Gaulish in France. But later,
when Celtic tribes occupied Ireland, Northern England, Wales, their tongues
were divided according to tribal divisions. These languages will later become
Welsh, Irish Gaelic, Cornish, but from that time no signs remained, because the
Celts did not invent writing yet. Not much is left from Celtic languages in
English. Though many place names and names for rivers are surely Celtic (like
Usk - from Celtic *usce "water", or Avon - from *awin
"river"), the morphology and phonetics are untouched by the Celtic
influence. Some linguists state that the word down comes from Celtic
*dún "down"; other examples of Celtic influence in place names
are tne following:
cothair (a fortress) - Carnarvon
uisge (water) - Exe, Usk, Esk
dun, dum (a hill) - Dumbarton, Dumfries, Dunedin
llan (church) - Llandaff, Llandovery, Llandudno
coil (forest) - Kilbrook, Killiemore
kil (church) - Kilbride, Kilmacolm
ceann (cape) - Kebadre, Kingussie
inis (island) - Innisfail
inver (mountain) - Inverness, Inverurie
bail (house) - Ballantrae, Ballyshannon,
and, certainly, the word whiskey which means the same as Irish uisge
"water". But this borrowing took place much later.
In the 1st century AD first Roman colonists begin to
penetrate in Britain; Roman legions built roads, camps, founded towns and
castles. But still they did not manage to assimilate the Celts, maybe because
they lived apart from each other and did not mix. Tens of Latin words in
Britain together with many towns, places and hills named by Romans make up the
Roman heritage in the Old English. Such cities as Dorchester, Winchester,
Lancaster, words like camp, castra, many terms of the Christian religion and
several words denoting armaments were borrowed at that time by Britons, and
automatically were transferred into the Old English, or Anglo-Saxon language
already when there was no Romans in the country.
In 449 the legendary leaders of two Germanic tribes,
Hengist and Horsa, achieved British shores on their ships. The Anglo-Saxon
conquest, however, lasted for several centuries, and all this period Celtic
aborigines moved farther and farther to the west of the island until they manage
to fortify in mountainous Wales, in Corwall, and preserved their kingdoms in
Scotland. Germanic tribes killed Celtic population, destroyed Celtic and former
Roman towns and roads. In the 5th century such cities as Durovern in Kent,
Virocon, Trimontii, Camulodunum, were abandoned by the population.
Angles settled around the present-day Noridge, and in
Northern England; Saxons, the most numerous of the tribes, occupied all Central
England, the south of the island and settled in London (Londinii at that time).
Jutes and Frises, who probably came to Britain a bit later, settled on the
island of White and in what is now Kent - the word Kent derives from the name
of the Celtic tribe Cantii. Soon all these tribes founded their separate
kingdoms, which was united after centuries of struggle only in 878 by Alfred,
king of Wessex. Before that each of the tribes spoke its language, they were
similar to each other but had differences which later became the dialectal
peculiarities of Old English.
Now a little bit about the foreign influence in Old
English. From the 6th century Christianity start activities in Britain, the
Bible is translated into Old English, and quite a lot of terms are borrowed
from Latin at that time: many bishops, missionaries and Pope's officials come
from Rome. The next group of foreign loanwords were taken from Scandinavian
dialects, after the Vikings occupied much of the country in the 9th - 11th
centuries. Scandinavian languages were close relatives with Old English, so the
mutual influence was strong enough to develop also the Old English morphology,
strengthening its analytic processes. Many words in the language were either
changed to sound more Scandinavian, or borrowed.
The Old English language, which has quite a lot of
literature monuments, came to the end after the Norman conquest in 1066. The
new period was called Middle English.
The Old English Substantive.
The substantive in Indo-European has always three main
categories which change its forms: the number, the case, the gender. It ias
known that the general trend of the Indo-European family is to decrease the
number of numbers, cases and genders from the Proto-Indo-European stage to
modern languages. Some groups are more conservative and therefore keep many
forms, preserving archaic language traits; some are more progressive and lose
forms or transform them very quickly. The Old English language, as well as
practically all Germanic tongues, is not conservative at all: it generated
quite a lot of analytic forms instead of older inflections, and lost many other
of them.
Of eight Proto-Indo-European cases, Old English keeps
just four which were inherited from the Common Germanic language. In fact,
several of original Indo-European noun cases were weak enough to be lost
practically in all branches of the family, coinciding with other, stronger
cases. The ablative case often was assimilated by the genitive (in Greek,
Slavic, Baltic, and Germanic), locative usually merged with dative (Italic,
Celtic, Greek), and so did the instrumental case. That is how four cases
appeared in Germanic and later in Old English - nominative, genitive,
accusative and dative. These four were the most ancient and therefore stable in
the system of the Indo-European morphology.
The problem of the Old English instrumental case is
rather strange - this case arises quite all of a sudden among Germanic tongues
and in some forms is used quite regularly (like in demonstrative pronouns). In
Gothic the traces of instrumental and locative though can be found, but are
considered as not more than relics. But the Old English must have
"recalled" this archaic instrumental, which existed, however, not for
too long and disappeared already in the 10th century, even before the Norman
conquest and transformation of the English language into its Middle stage.
As for other cases, here is a little pattern of their usage
in the Old English syntax.
1. Genitive - expresses the possessive
menaing: whose? of what?
Also after the expression meaning
full of , free of , worthy of , guilty of, etc.
2. Dative - expresses the object
towards which the action is directed.
After the after the verbs like
"say to smb", "send smb", "give to smb";
"known to smb", "necessary for smth / smb", "close to
smb", "peculiar for smth".
Also in the expressions
like from the enemy, against the wind, on the shore.
3. Accusative - expresses the
object immediately affected by the action (what?), the direct object.
Three genders were strong enough, and only northern
dialects could sometimes lose their distinction. But in fact the lose of
genders in Middle English happened due to the drop of the case inflections,
when words could no longer be distinguished by its endings. As for the numbers,
the Old English noun completely lost the dual, which was preserved only in
personal pronouns (see later).
All Old English nouns were divided into strong and
weak ones, the same as verbs in Germanic. While the first had a branched
declension, special endings for different numbers and cases, the weak
declension was represented by nouns which were already starting to lose their
declension system. The majority of noun stems in Old English should be referred
to the strong type. Here are the tables for each stems with some comments - the
best way of explaining the grammar.
a-stems
Singular
Nom. stán (stone) scip (ship)
bán (bone) reced (house) níeten (ox)
Gen.
stánes
scipes
bánes
recedes
níetenes
Dat.
stáne
scipe
báne
recede
níetene
Acc.
stán
scip
bán
reced
níeten
Plural
Nom. stánas
scipu
bán
reced
níetenu
Gen.
stána scipa
bána
receda
níetena
Dat. stánum
scipum
bánum
recedum níetenum
Acc. stánas
scipu
bán
reced
níetenu
This type of stems derived from masculine and neuter noun o-stems
in Proto-Indo-European. First when I started studying Old English I was
irritated all the time because I couldn't get why normal Indo-European o-stems
are called a-stems in all books on Old English. I found it a
silly and unforgivable mistake until I understood that in Germanic the
Indo-European short o became a, and therefore the
stem marker was also changed the same way. So the first word here, stán,
is masculine, the rest are neuter. The only difference in declension is the
plural nominative-accusative, where neuter words lost their endings or have -u,
while masculine preserved -as.
A little peculiarity of those words who have the sound [æ] in the stem
and say farewell to it in the plural:
Masculine
Neuter
Sing.
Pl.
Sing.
Pl.
N dæg (day) dagas
fæt (vessel) fatu
G dæges
daga
fætes fata
D dæge
dagum
fæte
fatum
A dæg dagas
fæt
fatu
Examples of a-stems:
earm (an arm), eorl, helm (a helmet),
hring (a ring), múþ (a mouth); neuter
ones - dor (a gate), hof (a courtyard), geoc
(a yoke), word, déor (an animal), bearn
(a child), géar (a year).
ja-stems
Singular
Masculine
Neuter
N hrycg (back) here (army) ende (end) cynn (kind)
ríce (realm)
G hrycges
heriges
endes
cynnes ríces
D
hrycge herige
ende
cynne ríce
A
hrycg
here
ende
cynn
ríce
Plural
N hrycgeas
herigeas
endas
cynn
ríciu
G hrycgea
herigea
enda
cynna rícea
D hrycgium
herigum
endum
cynnum rícium
A hrycgeas
herigeas
endas
cynn
ríciu
Again the descendant of
Indo-European jo-stem type, known only in masculine and neuter.
In fact it is a subbranch of o-stems, complicated by the i
before the ending: like Latin lupus and filius.
Examples of this type: masculine - wecg (a wedge), bócere
(a scholar), fiscere (a fisher); neuter - net, bed,
wíte (a punishment).
wa-stems
Singular
Plural
Masc.
Neut.
Masc. Neut.
N bearu (wood) bealu (evil)
bearwas bealu (-o)
G bearwes
bealwes
bearwa bealwa
D bearwe
bealwe
bearwum bealwum
A bearu (-o) bealu
(-o) bearwas bealu (-o)
Just to mention. This is one more peculiarity of good old a-stems
with the touch of w in declension. Interesting that the majority
of this kind of stems make abstract nouns. Examples: masculine - snáw
(snow), þéaw (a custom); neuter - searu
(armour), tréow (a tree), cnéw (a
knee)
ó-stems
Sg.
N swaþu (trace) fór (journey) tigol (brick)
G
swaþe
fóre
tigole
D
swaþe
fóre
tigole
A
swaþe
fóre
tigole
Pl.
N swaþa
fóra
tigola
G swaþa
fóra
tigola
D swaþum
fórum
tigolum
A swaþa
fóra
tigola
Another major
group of Old English nouns consists only of feminine nouns. Funny but in
Indo-European they are called a-stems. But Germanic turned vowels
sometimes upside down, and this long a became long o.
However, practically no word of this type ends in -o, which was
lost or transformed. The special variants of ó-stems are jo-
and wo-stems which have practically the same declension
but with the corresponding sounds between the root and the ending.
Examples of ó-stems: caru (care), sceamu
(shame), onswaru (worry), lufu (love), lár
(an instruction), sorg (sorrow), þrág
(a season), ides (a woman).
Examples of jó-stems: sibb (peace), ecg
(a blade), secg (a sword), hild (a fight), æx
(an axe).
Examples of wó-stems: beadu (a battle),
nearu (need), læs (a beam).
i-stems
Masc.
Neut.
Sg.
N sige (victory) hyll (hill) sife (sieve)
G
siges
hylles sifes
D
sige
hylle sife
A
sige
hyll sife
Pl.
N sigeas
hyllas sifu
G sigea
hylla sifa
D sigum
hyllum sifum
A sigeas
hyllas sifu
The tribes and nations were usually of this
very type, and were used always in plural: Engle (the Angles), Seaxe
(the Saxons), Mierce (the Mercians), Norþymbre
(the Northumbrians), Dene (the Danish)
N Dene
G Dena (Miercna, Seaxna)
D Denum
A Dene
Fem.
Sg. Pl.
N hyd (hide) hýde, hýda
G hýde
hýda
D hýde
hýdum
A
hýd
hýde, hýda
This kind of stems included all three genders
and derived from the same type of Indo-European stems, frequent also in other branches
and languages of the family.
Examples: masculine - mere (a sea), mete (food),
dæl (a part), giest (a guest), drync
(a drink); neuter - spere (a spear); feminine - cwén
(a woman), wiht (a thing).
u-stems
Masc.
Fem.
Sg.
N sunu (son)feld (field) duru (door) hand (hand)
G suna
felda
dura handa
D suna
felda
dura handa
A sunu
feld
duru hand
Pl.
N suna
felda
dura handa
G suna
felda
dura handa
D sunum
feldum
durum handum
A suna
felda
dura handa
They
can be either masculine or feminine. Here it is seen clearly how Old English
lost its final -s in endings: Gothic had sunus and handus,
while Old English has already sunu and hand
respectively. Interesting that dropping final consonants is also a general
trend of almost all Indo-European languages. Ancient tongues still keep them
everywhere - Greek, Latin, Gothic, Old Prussian, Sanskrit, Old Irish; but
later, no matter where a language is situated and what processes it undergoes,
final consonants (namely -s, -t, often -m, -n) disappear,
remaining nowadays only in the two Baltic languages and in New Greek.
Examples: masculine - wudu (wood), medu
(honey), weald (forest), sumor (a summer); fem. - nosu
(a nose), flór (a floor).
The
other type of nouns according to their declension was the group of Weak nouns,
derived from n-nouns is Common Germanic. Their declension is
simple and stable, having special endings:
Masc.
Fem.
Neut.
Sg.
N nama (name) cwene (woman) éage (eye)
G naman
cwenan
éagan
D naman
cwenan
éagan
A naman
cwenan
éage
Pl.
N naman
cwenan
éagan
G namena
cwenena
éagena
D namum
cwenum
éagum
A naman
cwenan
éagan
Examples: masc. - guma (a man), wita (a wizard),
steorra (a star), móna (the Moon), déma
(a judge); fem. - eorþe (Earth), heorte (a
heart), sunne (Sun); neut. - éare (an ear).
And now the last
one which is interesting due to its special Germanic structure. I am speaking
about the root-stems which according to Germanic laws of Ablaut, change the
root vowel during the declension. In Modern English such words still exist, and
we all know them: goose - geese, tooth - teeth, foot - feet, mouse - mice
etc. At school they were a nightmare for me, now they are an Old English
grammar. Besides, in Old English time they were far more numerous in the
language.
Masc.
Fem.
Sg.
N mann fót (foot) tóþ
(tooth) | hnutu (nut) bóc (book) gós
(goose) mús (mouse) burg (burg)
G mannes fótes
tóþes |
hnute
bóce
góse
múse
burge
D menn
fét
téþ
| hnyte
béc
gés
mýs
byrig
A mann
fót
tóþ
| hnutu
bók
gós
mús
burg
Pl.
N menn
fét
téþ
| hnyte
béc
gés
mýs
byrig
G manna
fóta
tóþa
| hnuta bóca
gósa
músa
burga
D mannum fótum
tóþum |
hnutum
bócum gósum
músum burgum
A menn
fét
téþ
| hnyte
béc
gés
mýs
byrig
The general rule is the so-called i-mutation,
which changes the vowel. The conversion table looks as follows and never fails
- it is universally right both for verbs and nouns. The table of i-mutation
changes remains above.
Examples: fem. - wífman (a woman), ác
(an oak), gát (a goat), bróc
(breeches), wlóh (seam), dung (a dungeon), furh
(a furrow), sulh (a plough), grut (gruel), lús
(a louse), þrul (a basket), éa
(water), niht (a night), mæ'gþ (a
girl), scrúd (clothes).
There are still some other types of declension, but
not too important fro understanding the general image. For example, r-stems
denoted the family relatives (dohtor 'a daughter', módor
'a mother' and several others), es-stems usually meant children
and cubs (cild 'a child', cealf 'a calf'). The most
intriguing question that arises from the picture of the Old English declension
is "How to define which words is which kind of stems?". I am sure you
are always thinking of this question, the same as I thought myself when first
studying Old English. The answer is "I don't know"; because of the
loss of many endings all genders, all stems and therefore all nouns mixed in
the language, and one has just to learn how to decline this or that word. This
mixture was the decisive step of the following transfer of English to the
analytic language - when endings are not used, people forget genders and cases.
In any solid dictionary you will be given a noun with its gender and kind of
stem. But in general, the declension is similar for all stems. One of the most
stable differences of masculine and feminine is the -es (masc.)
or -e in genitive singular of the Strong declension.
Now I am giving another table,
the general declension system of Old English nouns. Here '-' means a zero
ending.
Strong declension (a,
ja, wa, у, jу, wу, i -stems).
|
Masculine
|
Neutral
|
Feminine
|
|
Singular
|
Plural
|
Singular
|
Plural
|
Singular
|
Plural
|
Nominative
|
-
|
-as
|
-
|
-u (-)
|
-
|
Genitive
|
-es
|
-a
|
-es
|
-a
|
-e
|
-a
|
Dative
|
-e
|
-um
|
-e
|
-um
|
-e
|
-um
|
Accustive
|
-
|
-as
|
-
|
-u (-)
|
-e
|
-a
|
|
Weak
declension
|
u-stems
|
|
Singular
|
Plural
|
Singular
|
Plural
|
Nominative
|
-
|
-an
|
-
|
-a
|
Genitive
|
-an
|
-ena
|
-a
|
-a
|
Dative
|
-an
|
-um
|
-a
|
-um
|
Accustive
|
-an
|
-an
|
-
|
-a
|
The Old English Adjective.
In all historical Indo-European languages adjectives
possess practically the same morphological features as the nouns, the the
sequence of these two parts of speech is an ordinary thing in Indo-European.
However, the Nostratic theory (the one which unites Altaic, Uralic, Semitic,
Dravidian and Indo-European language families into one Nostratic super-family,
once speaking a common Proto-Nostratic language) represented by Illych-Svitych
and many other famous linguists, states that adjectives in this Proto-Nostratic
tongue were morphologically closer to the verbs than to the nouns.
This theory is quite interesting, because even in
Proto-Indo-European, a language which was spoken much later than
Proto-Nostratic, there are some proofs of the former predicative function of
the adjectives. In other families of the super-family this function is even
more clear. In Altaic languages, and also in Korean and Japanese, which
are originally Altaic, the adjective plays the part of the predicate, and in
Korean, for example, the majority of adjectives are predicative. It means that
though they always denote the quality of the noun, they act the same way as
verbs which denote action. Adjective "red" is actually translated
from Japanese as "to be red", and the sentence Bara-wa utsukusii
will mean "the rose is beautiful", while bara is "a
rose", -wa is the nominative marker, and utsukusii is
"to be beautiful". So no verb here, and the adjective is a predicate.
This structure is typical for many Altaic languages, and probably was normal
for Proto-Nostratic as well.
The Proto-Indo-European language gives us some stems
which are hard to denote whether they used to mean an adjective or a verb. Some
later branches reflect such stems as verbs, but other made them adjectives. So
it was the Proto-Indo-European epoch where adjectives as the part of speech
began to transform from a verbal one to a nominal one. And all Indo-European
branches already show the close similarity of the structure of adjectives and
nouns in the language. So does the Old English language, where adjective is one
of the nominal parts of speech.
As
well as the noun, the adjective can be declined in case, gender and number.
Moreover, the instrumental case which was discussed before was preserved in
adjectives much stronger than in nouns. Adjectives must follow sequence with
nouns which they define - thet is why the same adjective can be masculine,
neuter and feminine and therefore be declined in two different types: one for
masculine and neuter, the other for feminine nouns. The declension is more or
less simple, it looks much like the nominal system of declension, though there
are several important differences. Interesting to know that one-syllable
adjectives ("monosyllabic") have different declension than
two-syllable ones ("disyllabic"). See for yourselves:
Strong Declension
a, ó-stems
Monosyllabic
Sg.
Masc.
Neut. Fem.
N blæc (black) blæc blacu
G blaces
blaces blæcre
D blacum
blacum blæcre
A blæcne
blæc blace
I blace
blace -
Pl.
N blace
blacu blaca
G blacra
blacra blacra
D blacum
blacum blacum
A blace
blacu blaca
Here "I" means that very instrumental case,
answering the question (by what? with whom? with the help of what?).
Disyllabic
Masc.
Neut.
Fem.
Sg.
N éadig (happy)
éadig éadigu
G
éadiges
éadiges éadigre
D éadigum
éadigum éadigre
A
éadigne
éadig éadige
I
éadige
éadige
Pl.
N
éadige
éadigu éadiga
G
éadigra
éadigra éadigra
D éadigum
éadigum éadigum
A
éadige
éadigu éadigu
So not many new endings: for accusative singular we
have -ne, and for genitive plural -ra, which cannot
be met in the declension of nouns. The difference between monosyllabic and
disyllabic is the accusative plural feminine ending -a / -u.
That's all.
ja, jó-stems (swéte
- sweet)
Sg.
Pl.
Masc.
Neut.
Fem. Masc.
Neut. Fem.
N swéte
swéte swétu
swéte swétu
swéta
G swétes
swétes swétre
swétra swétra swétra
D swétum swétum
swétre swétum swétum
swétum
A swétne
swéte swéte
swéte swétu
swéta
I swéte
swéte -
wa, wó-stems
Sg.
Masc.
Neut.
Fem.
N nearu (narrow)
nearu nearu
G nearwes
nearwes nearore
D nearwum
nearwum nearore
A
nearone
nearu nearwe
I
nearwe
nearwe
Pl.
N nearwe
nearu nearwa
G nearora
nearora nearora
D nearwum
nearwum nearwum
A
nearwe
nearu nearwa
Actually, some can just omit all those examples - the
adjectival declension is the same as a whole for all stems, as concerns the
strong type. In general, the endings look the following way, with very few
varieties (note that "-" means the null ending):
As for weak adjectives, they also exist in the
language. The thing is that one need not learn by heart which adjective is
which type - strong or weak, as you should do with the nouns. If you have a
weak noun as a subject, its attributive adjective will be weak as well. So - a
strong adjective for a strong noun, a weak adjective for a weak noun, the rule
is as simple as that.
Thus
if you say "a black tree" that will be blæc tréow (strong),
and "a black eye" will sound blace éage. Here is
the weak declension example (blaca - black):
Sg.
Pl.
Masc.
Neut. Fem.
N blaca
blace
blace blacan
G blacan
blacan
blacan blæcra
D blacan
blacan
blacan blacum
A blacan
blace
blacan blacan
Weak declension has a single
plural for all genders, which is pleasant for those who don't want to remeber
too many forms. In general, the weak declension is much easier.
The last thing to be said about
the adjectives is the degrees of comparison. Again, the traditional
Indo-European structure is preserved here: three degrees (absolutive,
comparative, superlative) - though some languages also had the so-called
"equalitative" grade; the special suffices for forming comparatives
and absolutives; suppletive stems for several certain adjectives.
The suffices we are used to see
in Modern English, those -er and -est in weak,
weaker, the weakest, are the direct descendants of the Old English ones. At
that time they sounded as -ra and -est. See the
examples:
earm (poor) - earmra - earmost
blæc (black) - blæcra - blacost
eald (old) - ieldra - ieldest
strong - strengra - strengest
long - lengra - lengest
geong (young) - gingra - gingest
The most widespread and widely
used adjectives always had their degrees formed from another stem, which is
called "suppletive" in linguistics. Many of them are still seen in
today's English:
gód (good) - betera - betst
(or sélra - sélest)
yfel (bad) - wiersa - wierest
micel (much) - mára - máést
lýtel (little) - læ'ssa - læ'st
fear (far) - fierra - fierrest, fyrrest
néah (near) - néarra - níehst,
nýhst
æ'r (early) - æ'rra - æ'rest
fore (before) - furþra - fyrest (first)
Now you see what the word
"first" means - just the superlative degree from the adjective
"before, forward". The same is with níehst from néah
(near) which is now "next".
Old English affixation for
adjectives:
1. -ede (group "adjective stem +
substantive stem") - micelhéafdede (large-headed)
2. -ihte (from substantives with mutation) - þirnihte
(thorny)
3. -ig (from substantives with mutation) - hálig
(holy), mistig (misty)
4. -en, -in (with mutation) - gylden
(golden), wyllen (wóllen)
5. -isc (nationality) - Englisc, Welisc, mennisc
(human)
6. -sum (from stems of verbs, adjectives, substantives) - sibbsum
(peaceful), híersum (obedient)
7. -feald (from stems of numerals, adjectives) - þríefeald
(threefold)
8. -full (from abstract substantive stems) - sorgfull
(sorrowful)
9. -léás (from verbal and nominal stems) - slæpléás
(sleepless)
10. -líc (from substantive and adjective stems) - eorþlíc
(earthly)
11. -weard (from adjective, substantive, adverb stems) - inneweard
(internal), hámweard (homeward)
The
Old English Pronoun.
Pronouns were the only part of speech in Old English which preserved the
dual number in declension, but only this makes them more archaic than the rest
parts of speech. Most of pronouns are declined in numnber, case and gender, in
plural the majority have only one form for all genders.
We will touch each group of Old English pronouns and comment on them.
1.Personal pronouns
Through the last 1500 years mín
became mine, gé turned into you (ye as
a colloquial variant). But changes are still significant: the 2nd person
singular pronouns disappeared from the language, remaining only in poetic
speech and in some dialects in the north of England. This is really a strange
feature - I can hardly recall any other Indo-European language which lacks the
special pronoun for the 2nd person singular (French tu, German du,
Russian ty etc.). The polite form replaced the colloquial one,
maybe due to the English traditional "ladies and gentlemen" customs.
Another extreme exists in Irish Gaelic, which has no polite form of personal pronoun,
and you turn to your close friend the same way as you spoke with a prime
minister - the familiar word, translated into French as tu. It
can sound normal for English, but really funny for Slavic, Baltic, German
people who make a thorough distinction between speaking to a friend and to a
stranger
2. Demonstrative pronouns ('I' means the instrumental case)
3. Interrogative pronouns
N hwá hwæt
G hwæs hwæs
D hwæ'm hwæ'm
A hwone hwæt
I
- hwý,
hwí
These pronouns, which actually
mean the masculine and the neuter varieties of the same pronoun, derive from
Proto-Indo-European *kwis, with *kw becoming hw
in Germanic languages. In Gothic the combination hw was
considered as one sound which is another proof that the Indo-European the
labiovelar sound kw was a single sound with some specific
articulation.
Later Germanic languages changed
the sound in a different way: in Norwegian it remained as hv, in
German turned into w (as in wer 'who', was 'what'),
in English finally changed into wh pronounced in most cases [w],
but somewhere also like [h] or [hw].
Interesting that the instrumental of the word hwæt,
once being a pronoun form, later became the word why in English. So
'why?' is originally an instrumental case of the interrogative pronoun.
Other interrogative pronouns, or
adverbs, as they are sometimes called, include the following, all beginning
with hw:
hwilc 'which?' - is declined as the strong
adjective (see adjectives
above)
hwonne 'when?' - this and following are not declined, naturally
hwæ'r 'where?'
hwider 'whither?'
hwonan 'whence?'
4. Other kinds of pronouns
They include definite, indefinite, negative and relative, all typical for
Indo-European languages. All of them still exist in Modern English, and all of
them are given here:
a) definite
gehwá (every) - declined the same way as hwá
gehwilc (each),
ægþer (either),
æ'lc (each),
swilc (such) - all declined like strong adjectives
sé ylca (the same) - declined like a weak adjective
b) indefinite
sum (some),
æ'nig (any) - both behave the same way as strong
adjectives
c) negative
nán, næ'nig (no, none) - declined like strong
adjectives
d) relative
þe (which, that)
séþe (which, that) - they are not declined
In Proto-Indo-European and in many ancient Indo-European
languages there was a special kind of declension calleed pronominal, using only
by pronouns and opposed to the one used by nouns, adjectives and numerals. Old
English lost it, and its pronouns use all the same endings as the nouns and
adjectives. Maybe the only inflection which remembers the Proto-language times,
is the neuter nominative -t in hwæt and þæt,
the ancient ending for inanimate (inactive) nouns and pronouns.
The Old English Numeral.
It is obvious that all Indo-European languages have the general
trend of transformation
from the synthetic (or inflectional) stage to the analytic
one. At least for the latest 1,000 years this trend could be observed in all
branches of the family. The level of this analitization process in each single
language can be estimated by several features, their presence or absence in the
language. One of them is for sure the declension of the numerals. In
Proto-Indo-European all numerals, both cardinal and ordinal, were declined, as
they derived on a very ancient stage from nouns or adjectives, originally being
a declined part of speech. There are still language groups within the family
with decline their numerals: among them, Slavic and Baltic are the most typical
samples. They practically did not suffer any influence of the analytic
processes. But all other groups seem to have been influenced somehow. Ancient
Italic and Hellenic languages left the declension only for the first four
cardinal pronouns (from 1 to 4), the same with ancient Celtic.
The Old English language
preserves this system of declension only for three numerals. It is therefore
much easier to learn, though not for English speakers I guess - Modern English
lacks declension at all.
Here is the list of the cardinal numerals:
Ordinal numerals use the suffix -ta or -þa,
etymologically a common Indo-European one (*-to-).
The
Old English Adverb.
Adverbs can be either primary (original adverbs)
or derive from the adjectives. In fact, adverbs appeared in the language rather
late, and eraly Proto-Indo-European did not use them, but later some auxiliary
nouns and pronouns losing their declension started to play the role of
adverbial modifiers. That's how thew primary adverbs emerged.
In Old English the basic primary adverbs were the following ones:
þa (then)
þonne (then)
þæ'r (there)
þider (thither)
nú (now)
hér (here)
hider (hither)
heonan (hence)
sóna (soon)
oft (often)
eft (again)
swá (so)
hwílum (sometimes).
Secondary adverbs originated from the
instrumental singular of the neuter adjectives of strong declension. They all
add the suffix -e: wide (widely), déope
(deeply), fæste (fast), hearde (hard).
Another major sugroup of them used the suffixes -líc, -líce
from more complexed adjectives: bealdlíce (boldly), freondlíce
(in a friendly way).
Adverbs, as well as adjectives, had their degrees of comparison:
wíde - wídor - wídost (widely
- more widely - most widely)
long - leng (long - longer)
feorr (far) - fierr
sófte (softly) - séft
éaþe (easily) - íeþ
wel (well) - betre - best
yfele (badly) - wiers, wyrs - wierst
micele (much) - máre - mæ'st
The Old English Verb.
Old English system had strong and
weak verbs: the ones which used the ancient Germanic type of conjugation (the
Ablaut), and the ones which just added endings to their past and participle
forms. Strong verbs make the clear majority. According to the traditional
division, which is taken form Gothic and is accepted by modern linguistics, all
strong verbs are distinguished between seven classes, each having its
peculiarities in conjugation and in the stem structure. It is easy to define
which verb is which class, so you will not swear trying to identify the type of
conjugation of this or that verb (unlike the situation with the substantives).
Here is the table which is composed for you to see the root
vowels of all strong verb classes. Except the VII class, they all have exact
stem vowels for all four main forms:
Now let us see what Old English strong verbs of
all those seven classes looked like and what were their main four forms. I
should mention that besides the vowel changes in the stem, verbal forms also
changed stem consonants very often. The rule of such changes is not mentioned
practically in any books on the Old English language, though there is some. See
for yourselves this little chart where the samples of strong verb classes are
given with their four forms:
Infinitive, Past singular, Past plural, Participle II (or Past Participle)
Class I
wrítan (to write), wrát, writon, writen
snípan (to cut), snáþ, snidon, sniden
Other examples: belífan (stay), clífan
(cling), ygrípan (clutch), bítan
(bite), slítan (slit), besmítan
(dirty), gewítan (go), blícan
(glitter), sícan (sigh), stígan (mount),
scínan (shine), árísan
(arise), líþan (go).
Class II
béodan (to offer), béad, budon, boden
céosan (to choose), céas, curon, coren
Other examples: créopan (creep), cléofan
(cleave), fléotan (fleet), géotan
(pour), gréotan (weep), néotan
(enjoy), scéotan (shoot), léogan
(lie), bréowan (brew), dréosan
(fall), fréosan (freeze), forléosan
(lose).
Class III
III a) a nasal consonant
drincan (to drink), dranc, druncon, druncen
Other: swindan (vanish), onginnan
(begin), sinnan (reflect), winnan (work), gelimpan
(happen), swimman (swim).
III b) l + a consonant
helpan (to help), healp, hulpon, holpen
Other: delfan (delve), swelgan
(swallow), sweltan (die), bellan (bark), melcan
(milk).
III c) r, h + a consonant
steorfan (to die), stearf, sturfon, storfen
weorþan (to become), wearþ, wurdon, worden
feohtan (to fight), feaht, fuhton, fohten
More: ceorfan (carve), hweorfan
(turn), weorpan (throw), beorgan (conceal), beorcan
(bark).
Class IV
stelan (to steal), stæ'l, stæ'lon, stolen
beran (to bear), bæ'r, bæ'ron, boren
More: cwelan (die), helan
(conceal), teran (tear), brecan (break).
Class V
tredan (to tread), træ'd, træ'don, treden
cweþan (to say), cwæ'þ,
cwæ'don, cweden
More: metan (measure), swefan
(sleep), wefan (weave), sprecan (to speak), wrecan
(persecute), lesan (gather), etan (eat), wesan
(be).
Class VI
faran (to go), fór, fóron, faren
More: galan (sing), grafan
(dig), hladan (lade), wadan (walk), dragan
(drag), gnagan (gnaw), bacan (bake), scacan
(shake), wascan (wash).
Class VII
hátan (to call), hét, héton,
háten
feallan (to fall), feoll, feollon, feallen
cnéawan (to know), cnéow, cnéowon,
cnáwen
More: blondan (blend), ondræ'dan
(fear), lácan (jump), scadan (divide), fealdan
(fold), healdan (hold), sponnan (span), béatan
(beat), blówan (flourish), hlówan
(low), spówan (flourish), máwan
(mow), sáwan (sow), ráwan (turn).
So the rule from the table above is observed carefully. The VII class was
made especially for those verbs which did not fit into any of the six classes.
In fact the verbs of the VII class are irregular and cannot be explained by a
certain exact rule, though they are quite numerous in the language.
Examining verbs of Old English comparing to those
of Modern English it is easy to catch the point of transformation. Not only the
ending -an in the infinitive has dropped, but the stems were
subject to many changes some of which are not hard to find. For example, the
long í in the stem gives i with an open
syllable in the modern language (wrítan > write, scínan
> shine). The same can be said about a, which nowadays
is a in open syllables pronounced [æ] (hladan >
lade). The initial combination sc turns to sh;
the open e was transformed into ea practically
everywhere (sprecan > speak, tredan > tread, etc.). Such
laws of transformation which you can gather into a small table help to recreate
the Old word from a Modern English one in case you do not have a dictionary in
hand, and therefore are important for reconstruction of the languages.
Weak verbs in Old English (today's English regular verbs) were conjugated in
a simpler way than the strong ones, and did not use the ablaut interchanges of
the vowel stems. Weak verbs are divided into three classes which had only
slight differences though. They did have the three forms - the infinitive, the
past tense, the participle II. Here is the table.
Class I
Regular verbs
Inf.
Past PP
déman (to judge), démde, démed
híeran (to hear), híerde, híered
nerian (to save), nerede, nered
styrian (to stir), styrede, styred
fremman (to commit), fremede, fremed
cnyssan (to push), cnysede, cnysed
When the suffix is preceded by a voiceless consonant the ending
changes a little bit:
cépan (to keep), cépte, cépt /
céped
grétan (to greet), grétte, grét /
gréted
If the verb stem ends in consonant plus d or t:
sendan (to send), sende, send / sended
restan (to rest), reste, rest / rested
Irregular
sellan (to give), sealde, seald
tellan (to tell), tealde, teald
cwellan (to kill), cwealde, cweald
tæ'can (to teach), táhte, táht
ræ'can (to reach), ráhte, ráht
bycgan (to buy), bohte, boht
sécan (to seek), sóhte, sóht
wyrcan (to work), worhte, worht
þencan (to think), þóhte,
þóht
bringan (to bring), bróhte, bróht
Other examples of the I class weak verbs just for your interest: berian
(beat), derian (harm), erian (plough), ferian
(go), herian (praise), gremman (be angry), wennan
(accustom), clynnan (sound), dynnan (resound), hlynnan
(roar), hrissan (tremble), sceþþan
(harm), wecgean (move), féran (go),
læ'ran (teach), dræfan (drive), fýsan
(hurry), drýgean (dry), híepan
(heap), métan (to meet), wýscean
(wish), byldan (build), wendan (turn), efstan
(hurry). All these are regular.
Class II
macian (to make), macode, macod
lufian (to love), lufode, lufod
hopian (to hope), hopode, hopod
Tis class makes quite a small group of verbs, all of them having -o-
before the past endings. Other samples: lofian (praise), stician
(pierce), eardian (dwell), scéawian (look), weorþian
(honour), wundrian (wonder), fæstnian
(fasten), mærsian (glorify).
Class III
habban (to have), hæfde, hæfd
libban (to live), lifde, lifd
secgan (to say), sægde, sægd
hycgan (to think), hogde, hogod
þréagan (to threaten), þréade,
þréad
sméagan (to think), sméade, sméad
fréogan (to free), fréode, fréod
féogan (to hate), féode, féod
Old English verbs are conjugated
having two tenses - the Present tense and the Past tense, and three moods -
indicative, subjunctive, and imperative. Of these, only the subjunctive mood
has disappeared in the English language, acquiring an analytic construction
instead of inflections; and the imperative mood has coincided with the
infinitive form (to write - write!). In the Old English period they all
looked different.
The common table of the verb
conjugation is given below. Here you should notice that the Present tense has
the conjugation for all three moods, while the Past tense - for only two moods
(no imperative in the Past tense, naturally). Some more explanation should be
given about the stem types.
In fact all verbal forms were
generated in Old English from three verb stems, and each verb had its own three
ones: the Infinitive stem, the Past Singular stem, the Past Plural stem. For
the verb wrítan, for example, those three stems are: wrít-
(infinitive without the ending -an), wrát-
(the Past singular), writ- (the Past plural without the ending -on).
The table below explains where to use this or that stem.
Additionally, the participles
(Participle I and Participle II) are formed by the suffix -ende
to the Infinitive stem (participle I), or the prefix ge- + the
Past Plural stem + the ending -en (Participle II).
Tired of the theory? Here is the preactice. We give several examples of the
typical verbs - first strong, then weak, then irregular.
Infinitive
Participle
wrítan
I wrítende II gewriten
Class II weak - lufian (to love)
Pres.
Past
Ind.
Subj.
Imp.
Ind. Subj.
Sg. 1
lufie
-
lufode
2 lufast
}lufie
lufa lufodest } lufode
3
lufaþ
-
lufode
Pl. lufiaþ lufien 2
lufiaþ lufodon lufoden
Part.
I lufiende II gelufod
Class III strong - bindan (to bind)
Pres.
Past
Ind.
Subj.
Imp.
Ind.
Subj.
Sg. 1
binde
- ¦ band, bond
2 bindest } binde
bind ¦
bunde } bunde
3
bindeþ
- ¦ band, bond
Pl. bindaþ binden
bindaþ ¦
bundon bunden
Inf. Part.
bindan
I bindende II gebunden
Class V strong - séon (to see)
Pres.
Past
Ind.
Subj.
Imp.
Ind. Subj.
Sg.1
séo
-
seah
2 síehst }
séo
seoh
sáwe } sáwe,
3
síehþ
-
seah sæge
Pl. séoþ
séon 2
séoþ
sawon sáwen
Participle
I séonde II gesewen, gesegen
Class VII strong - fón (to
catch)
Pres.
Past
Ind. Subj.
Imp.
Ind. Subj.
Sg. 1
fó
- feng
2 féhst }
fó
fóh
fenge } fenge
3
féhþ
- feng
Pl. fóþ
fón 2
fóþ
fengon fengen
Participle
I fónde II gefangen, gefongen
Class III weak - secgan (to say)
Pres.
Past
Ind.
Subj.
Imp.
Ind. Subj.
Sg.1
secge
-
sægde
2 sægst
}secge
sæge sægdest
}sægde
3
sægþ
-
sægde
Pl. secgaþ secgen 2
secgaþ sægdon sægden
Part.
I secgende II gesægd
Class III weak - libban (to live)
Pres.
Past
Ind.
Subj.
Imp.
Ind. Subj.
Sg.1
libbe
- lifde
2 liofast }libbe
liofa lifdest
} lifde
3
liofaþ
-
lifde
Pl. libbaþ libben 2
libbaþ lifdon lifden
Part.
I libbende II gelifd
A special group is made by the so-called
Present-Preterite verbs, which are conjugated combining two varieties of the
usual verb conjugation: strong and weak. These verbs, at all not more than
seven, are nowadays called modal verbs in English.
Present-Preterite verbs have their Present tense forms generated from the
Strong Past, and the Past tense, instead, looks like the Present Tense of the
Weak verbs. The verbs we present here are the following: witan
(to know), cunnan (can), þurfan (to need), dearan
(to dare), munan (to remember), sculan (shall), magan
(may).
Present of witan (= strong Past)
Ind.
Subj. Imp.
Sg. 1
wát
-
2 wast }
wite wite
3
wát
-
Pl. witon 2
witen witaþ
Past (= Weak)
Ind.
Subj.
Sg.1 wisse, wiste
2 wissest, wistest } wisse, wiste
3 wisse, wiste
Pl. wisson, wiston wissen,
wisten
Participles: I witende, II witen, gewiten
cunnan (can)
Pres.
Past
Ind.
Subj.
Ind. Subj.
Sg. 1
cann
cúþe
2 canst }
cunne cúþest }
cúþe
3
cann
cúþe
Pl. cunnon cunnen
cúþon cúþen
þurfan (need)
Sg. 1
þearf
þorfte
2 þearft }
þurfe þorftest }
þorfte
3
þearf
þorfte
Pl. þurfon
þurfen
þorfton þorften
magan (may)
Sg. 1
mæg
meahte mihte, mihten
2 meaht } mæge
meahtest
3
mæg
meahte
Pl. magon mægen
meahton
The main difference of verbs of this type in modern English is their expressing
modality, i.e. possibility, obligation, necessity. They do not require the
particle to before the infinitive which follows them. In Old English in
general no verb requires this particle before the infinitive. In fact, this to
before the infinitive form meant the preposition of direction.
And now finally a few irregular verbs, which used several different stems
for their tenses. These verbs are very important in Old English and are met
very often in the texts: wesan (to be), béon
(to be), gán (to go), dón (to do), willan
(will). Mind that there was no Future tense in the Old English
language, and the future action was expressed by the Present forms, just
sometimes using verbs of modality, willan (lit. "to wish to
do") or sculan (lit. "to have to do").
wesan (to be) - has got only the Present tense forms, uses the
verb béon in the Past
Present
Ind.
Subj. Imp.
Sg.1 eom -
2 eart } síe,
sý wes
3
is -
Pl. sind síen, sýn 2
wesaþ
béon (to be)
Present
Ind.
Subj. Imp.
Sg. 1
béo
-
2 bist
}béo béo
3
biþ
-
Pl. béoþ
béon 2 béoþ
Past
Ind.
Subj.
Sg. 1 wæs
2 wære } wære
3 wæs
Pl. wæron wæren
Participle I is béonde (being).
gán (to go)
Pres.
Past
Ind.
Subj. Imp.
Ind. Subj.
Sg.1
gá
- éode
2 gæ'st }
gá
gá
éodest } éode
3
gæ'þ
- éode
Pl. gáþ 2
gán gáþ
éodon éoden
Participles:
I gánde, gangende II gegán
So there were in fact two verbs meaning 'to be',
and both were colloquial. In Middle English, however, the verb wesan
replaced fully the forms of béon, and the words béo
(I am), bist (thou art) fell out of use. The Past tense forms was
and were are also derivatives from wesan.
Syntactically, the language had
only two main tenses - the Present and the Past. No progressive (or Continuous)
tenses were used, they were invented only in the Early Middle English period.
Such complex tenses as modern Future in the Past, Future Perfect Continuous did
not exist either. However, some analytic construction were in use, and first of
all the perfective constructions. The example Hie geweorc geworhten
hæfdon 'they have build a fortress' shows the exact
Perfect tense, but at that time it was not the tense really, just a participle
construction showing that the action has been done. Seldom you can also find
such Past constructions, which later became the Past Perfect Tense.
Verb syntax includes a number of
suffices and prefixes which can be met in Old English texts and especially in
poetry:
Suffices:
1. -s- (from substantive or adjective stems) - mæ'rsian
(to announce; from mæ're - famous)
2. -læc- - néálæcan (to
approach)
3. -ett- - bliccettan (to sparkle)
Prefixes
1. á- = out of, from - árísan
(arise), áwakan (awake), áberan
(sustain)
2. be- = over, around, by - begán (go
around), beþencan (think over), behéafdian
(behead)
3. for- = destruction or loss - fordón
(destroy), forweorþan (perish)
4. mis- = negation or bad quality - mislícian
(displease)
5. of- = reinfors - ofsléan (kill), oftéon
(take away)
6. on- = change or separation - onbindan
(unbind), onlúcan (unlock)
7. tó- = destruction - tóbrecan
(break)
The Old English Auxiliary Words.
These traditionally include prepositions, conjunctions, different particles and
interjections. All Indo-European languages have this system
of auxiliary parts of speech, though there are languages which lack some of
them. Japanese, for example, has no prepositions, and the service function in
the sentence belongs to postpositive words which have cases, the same as nouns.
Korean does not use any conjunctions, replacing them by about 50 different
kinds of verbal adverbs. As for Chinese, it simply does not make any
distinction in the sentence between basic and auxiliary words.
Most of Old English prepositions are easily recognizable:
Primary: of (of, out of), æt
(to), fram (from), tó (to), wiþ
(against), in, of, mid (with), on (on, at),
be (by, near, to, because of, about), þurh
(through), under, ofer (over), æfter
(after), bufan (above), út (out).
Secondary: beforan (before), bútan
(without), benorþan (north of), etc.
æt means 'to' and wiþ
means 'against'. In Germanic all prepositions divided into those who
used nouns in dative, accusative or genitive. But in the Old English period
this distinction begins to disappear, and only some of the prepositions use
dative (mid, bútan, sometimes on, in) or
genitive (fram, út, æfter).
Conjunctions included the following:
Primary: and / ond (and) , ac
(but), gif (if), or.
Secondary: ægþer ge... ge (both... and..., either ...
or...), hwonne (when), þa (when), þonne
(when), þéáh (though), þætte
(that), ær (before), swá... swá...
(so... as...).
And a few interjections: iá (yes), wá
(woe!, wow!), hwæt (there! what!).