Let there be inflections!
"To be or not to be" – that is a non-finite clause. And "to be" is an infinitive marked with
to.
Curiously enough, such
to-infinitives had a special inflectional ending back in the days of Old English. Here is an example:
(1) |
Forðæm |
swa |
hwa |
swa |
wilnað |
good |
to |
donne, |
|
for-that |
so |
who |
so |
wants |
good |
to |
do |
|
he |
wilnað |
good |
to |
habbanne |
& |
mid |
goode |
to |
bionne. |
|
he |
wants |
good |
to |
have |
and |
with |
good |
to |
be |
'Therefore whosoever desires to do good things, he desires to have
good things and to be with good things.' |
(coboeth,Bo:36.110.28.2177) |
Look at the infinitives in this sentence:
to donne 'to do,'
to habbanne 'to have,'
to bionne 'to be.'
See something strange? All these verbs end in a rather peculiar ending, namely
-ne. You’re looking at the
so-called Old English inflected infinitive.
Origin of the inflected infinitive
So, what is this strange ending? Where did it come from? Well, Old English had case markings for its nominal word classes,
like nouns (
house), determiners (
the) or adjectives (
old). One of them was the dative case. Dative
was used, among others, for indirect objects (like
Mary in
John offered Mary a job) and more generally for
discourse participants that are not in direct control of the action but are affected by it, like people who receive,
experience, profit or suffer from something. Dative endings were also required after some prepositions, most importantly,
after
to. For instance, in the Old English phrases
to Lundene 'to London' or
to heofene 'to heaven,'
the final vowel
-e is a dative marker.
Now, some of these
to prepositions plus a dative noun were quite remarkable because they behaved a
little bit like clauses. That means that the head noun after
to could have a direct relation with
another noun phrase. This other noun phrase often expressed someone who derives some advantage from the head noun,
a beneficiary or a profiteer. Look at the example below:
(2) |
þa |
þæt |
mæden |
wæs |
XV |
gear, |
|
when |
that |
maiden |
was |
fifteen |
years, |
|
þa |
wolde |
se |
fæder |
hi |
sellan |
sumum |
æþelon |
men |
to |
bryde.
|
|
then |
wanted |
the |
father |
her |
give |
some |
noble |
man |
to |
bride |
'When the girl was fifteen, her father wanted to marry her to a noble man.' |
(comart1,Mart_1_[Herzfeld-Kotzor]:De25,C.7.40) |
Translated literally, this sentence includes the expression 'some noble man to bride.' But what that really means
is that "some noble man" benefits from the "bride". The phrase could be translated as something like, the maiden
functioned "as a bride
for some noble man." There is an immediate relationship between the head noun,
"bride," and the noun phrase, "a noble man." Here is another example that makes the same point:
(3) |
We |
soðlice |
æfter |
ðeawlicum |
andgite |
cealf |
offriað |
Gode |
to |
lace. |
|
we |
truly |
after |
customary |
understanding |
calf |
offer |
God |
to |
sacrifice |
'As is the right tradition, we sacrifice a calf to God' |
(cocathom2,ÆCHom_II,_12.1:120.357.2629) |
As in the last example, there is a strange construction in this sentence, which literally reads,
'God to sacrifice.' What that means is that "God" is a profiteer of the sacrifice. The calf functions
"as a sacrifice
for God." As before, the head noun, "sacrifice," and the noun phrase, "God", have
a direct relationship.
What's more - the dependent noun phrase could not only be interpreted as a beneficiary. Instead, it could even
behave almost as if it was a direct object of the noun after
to. This is illustrated in the following sentence.
(4) |
Apollonius |
onfengc |
þam |
mædenne |
to |
lare |
|
Apollonius |
received |
the |
maiden |
to |
instruction |
'Apollonius received the girl in order to teach her' |
(coapollo,ApT:18.13.380) |
There is a complicated phrase in this example, "the maiden to instruction." Here, "the maiden" is not just an
entity that profits from "instruction." Instead, she is the goal of the teaching, she directly
receives instruction.
That's why one could translate the noun after
to
and the noun phrase before
to as if they were in a verb – object relation: "to teach the maiden."
But keep in mind, "the maiden"
is still dependent on another noun, "instruction, lore", not on a verb. Here is another illustration.
(5) |
Hasterbal, |
Hannibales |
broðor, |
for |
mid |
firde |
|
Hasdrubal, |
Hannibal’s |
brother, |
went |
with |
troop |
|
of |
Ispanium |
on |
Italie |
Hannibale |
to |
fultume. |
|
from |
Spain |
to |
Italy |
Hannibal |
to |
assistance. |
'Hasdrubal, Hannibal’s brother, came with an army to the aid of Hannibal from Spain to Italy.' |
(coorosiu,Or_4:10.105.27.2183) |
As in the previous sentence, an element before the preposition
to, the proper name "Hannibal," has a direct
relationship with the noun after the preposition
to, "help, assistance," which goes beyond a beneficiary
interpretation. "Hannibal" does not just profit from his brother's "help." Instead, he directly receives it; one
could say, he is almost like the object of "assistance." That's the reason why the phrase can be translated
as a verb - object structure: "in order to help Hannibal." But again, there is no verb after the preposition
to
here. The word "help, assistance" is clearly a noun.
However, since the preposition
to often occurred with these fascinating clause-like nouns, it was only a small step
to actually start using verbs in its place. The recipe is simple: instead of a noun, just stick any infinitive into the
slot after the preposition and treat it as if it were a noun. Ingenious! Nominalized non-finite forms like this are called
'gerunds' and like all nouns after the preposition
to, they had to be inflected for dative of course. That’s
where the inflectional ending on Old English infinitives came from: it’s an old marking of dative case. The following
example illustrates how this might have worked:
(6) |
Ic |
Ælfric |
munuc |
awende |
þas |
boc |
of |
ledenum |
bocum |
to |
engliscum |
gereorde |
|
I |
Aelfric |
monk |
turned |
this |
book |
from |
Latin |
books |
to |
English |
language |
|
þam |
mannum |
to |
rædenne |
þe |
þæt |
leden |
ne |
cunnon; |
|
the |
men |
to |
read |
who |
that |
Latin |
not |
know |
'I, monk Aelfric, translated this book from Latin books into the English language as
reading for those men who do not know Latin.'
|
(coprefcath2,ÆCHom_II_[Pref]:.4) |
Once again, there is an odd phrase in this sentence, literally translated as 'the men to read." As before, the
noun phrase before
to can be thought of as a beneficiary of the word after
to. That is to say,
"men who don't know Latin" derive an advantage from the "reading," or the English book functions "as a reading
for those men." The relations in this expression are therefore just as in the previous examples,
"as a bride for the noble man" and "as a sacrifice for God." There is only one
difference: the word after
to is not a noun but a verb-derived gerund now. And if you look closely, you'll see that
it does indeed show up with the inflectional dative ending
-ne.
(7) |
hy |
hraðe |
common |
hym |
to |
fultumigenne |
|
they |
quickly |
came |
them |
to |
help |
'They quickly came to their aid' |
(coaelhom,ÆHom_15:23.2149) |
In this example, there is a direct relationship between the pronoun "them" and "help" in the phrase, translated literally
as, 'them to help.' Not only is it possible to interpret "them" as the profiteer of "help," but even directly as its
recipient. In that sense, this example is similar to the object-like noun phrases in the above examples
"the maiden to instruction" and - in fact even exactly parallel to - "Hannibal to assistance." The crucial difference is,
once again, that the element after
to is no longer a noun but a form based on a verb. It is an infinitive inflected
for dative with the typical ending
-ne.
Voilà, the proto-form was born of what was later to become the Modern English
to-infinitive!
By Old English times, to-infinitives formed clauses
The state of affairs just described is very old.
The inflected infinitive exclusively functioned like a noun after
to only until Proto-Germanic times,
maybe until 300 A.D. or so. By the time Old English was written down, it already behaved like a verb in the
overwhelming majority of cases.
That means that the whole complex,
to plus the infinitive, developed from a prepositional phrase to a non-finite clause.
The word
to was recategorized from a preposition to a non-finite marker. And the infinitive itself changed from a
nominalized gerund to a true verbal form.
It's easy to see how this reanalysis might have happened. Take again the last example, literally translated as
'they quickly came them to help.'
Originally, the
to-infinitive was interpreted as a prepositional phrase,
"They quickly came
as help for them / in their support / for their assistance." The translations suggested for
to, "as", "in", "for", are prepositions. But look at the example again, 'they quickly came them to help.' Don't
you feel that it's perfectly possible to analyze this sentence differently? One could assume that "them" is really
the direct object of the infinitive "help", which would therefore be an actual verb. That's exactly what
speakers of prehistoric Old English must have done. They heard the sentence and assumed that it included an infinitival
clause, specifically, one that conveys the purpose of an action. They interpreted the sentence as "They quickly came
in order to help them / to support them / to assist them." The translations suggested for
to is thus the same non-finite marker,
to, as in Modern English infinitives.
The crucial difference between the interpretation of
to-infinitives as prepositional phrases or clauses is that the
latter but not the former contain their own, implicit subject.
For example, in the sentence
Democratic politicians promised to sanction the dictator,
it could be asked, "well, who do you understand to be the entity that would sanction the dictator?" The answer is
"Democratic politicians." So, "Democratic politicians" refers to the implicit subject of the
infinitival clause "to sanction the dictator." In contrast, the sentence
The economy created fewer jobs in response to more sanctions does not allow answering the question, "Well,
who do you understand to be the entitity that passed the sanctions?" We just don't know. The reason is that "to sanction the
dictator" is a clause and contains a subject while "to more sanctions" is a prepositional phrase and doesn't have a subject.
The table below might clarify the difference between the original and reanalyzed interpetations of the Old
English
to-infinitive.
|
Original interpretation |
Reanalyzed interpretation |
Translation: |
as a support for them |
to support them |
Category: |
a prepositional phrase |
a non-finite clause |
Status of to: |
a preposition |
a non-finite marker |
Status of infinitive: |
nominalized (gerund) |
verbal |
Crucial difference: |
doesn't have a subject |
has an implicit subject |
So, how do we know that the
to-infinitive already contained an implicit subject most of the time
even by Old English times?
Well, first of all, Old English
to-infinitives are often directly selected by main verbs. That means that
the whole complex,
to plus the infinitive plus other dependents, functions as the complement of the verb; it conveys
the content of the main clause verb phrase. Such an interpretation is only possible if the
to-infinitive heads a non-finite clause with its own, implicit subject. Here is an example of an Old English
to-infinitive that functions as a complement.
(8) |
hie |
wielnien |
to |
wietanne |
ðæt |
ðæt |
hie |
nyton. |
|
they |
desire |
to |
know |
that |
which |
they |
not-know |
'They desire to know that which they don't know.' |
(cocura,CP:30.203.6.1361) |
Clearly, the non-finite clause "to know that which they don't know" is the complement of "desire." The subject of the main
verb "desire" is also the subject of the infinitive "to know". In fact, this Old English example is totally parallel
to its Modern English translation -
hie wielnien to witanne is exactly like
they desire to know. The infinitive
has an implicit subject in both language stages.
Furthermore, there must be a subject in Old English
to-infinitive clauses because there are some examples where
the subject
has a tight semantic relationship with the infinitive, but not with the main verb. For example, the subject and the infinitive
may form a weather expression ("it's raining") or an idiom ("all hell broke loose"). Thus, when the subject shows up with
the main verb, it's still evident that it actually belongs to the infinitive (that's called a "raising" construction). Here
is an example:
(9) |
þær |
begann |
to |
brastligenne |
micel |
þunor. |
|
there |
began |
to |
rattle |
great |
thunder |
'a great thunder began to rattle.' |
(cocathom2,ÆCHom_II,_12.1:113.122.2464) |
The only entity that can "rattle" in this sentence, is "great thunder". In contrast, "great thunder" cannot literally "begin"
anything. Similarly, you cannot really say, "Great thunder wants / tries / hopes etc. something." Therefore, the subject
"great thunder" must belong to the
to-infinitive.
Finally, there are some cases where an infinitival clause with
to does in fact occur with an overt subject
- even though
this happens only very rarely. The following sentence is a case in point:
(10) |
do |
hit |
mon |
us |
to |
witanne, |
|
make |
it |
someone |
us |
to |
know |
'matters should be arranged in such a way that we will know it.' |
(cocura,CP:46.357.3.2412) |
This Old English sentence contains the non-finite clause "us to know it." It depends on the now outdated expression
"someone does" to mean "someone causes, makes, lets." The whole construction means "someone should cause the following:
that we know it." Hence, "us" functions as the subject of the
to-infinitive here.
All in all then, there can be
absolutely no doubt that by Old English times already, the
to-infinitive normally functioned as a real
clause and included its own subject. It behaved like a prepositional phrase in marginal contexts only.
The declension has left the language!
In twelfth century Middle English,
to-infinitives undergo further developments that turn them into an unambiguously
verbal category. They begin to be used as passives (e.g.
She was afraid [to be persecuted]), with perfective
have
(e.g.
The president seems [to have taken an unpopular position]), and include independent negation
(e.g.
Peter promised [to not do it again]).
Furthermore, there was a decline at the same time of the last vestiges
of
to-infinitives that were plausibly
analyzed as gerunds after the preposition
to and not as verbs. After all, in Modern English you cannot really say
things like "I translated you this book to read" anymore. Thus, learners of early Middle English were
presented with more and more evidence that the
to-infinitive should be treated like a verb and less and less evidence
that it could be used like a noun.
Add to this state of affairs the fact that case morphology eroded across the board
in early Middle English, and one might expect that the inflectional ending on infinitives would gradually
disappear. Just imagine the linguistic subconsciousness of a twelfth century Englishman, thinking,
"Why the heck is there a
-ne ending on my infinitives? It's a noun? What are you kidding me? I've never heard
the infinitive used like a noun. Ever! Dative? What? I don't even know what that is... You know what, I'm just gonna go
ahead and erase this erratic ending."
|
And this is exactly what happened! The Old English inflected infinitive died a fast
death in the late 11th and 12th centuries.
It is even possible to measure this decline empirically. The graph on the left-hand side
is based on data retrieved from Old and Middle English corpora.
It represents with red points the proportion of to-infinitves that have an inflectional ending
in dozens of Old and early Middle English texts. The size of the points is relative to the
number of examples in each text. The fitted blue line shows the general trend of
the individual data points - namely that inflected infinitives drop out of the language.
Click on the graphic to enlarge it.
You can learn more about this graph and the
data that was used to create it
on this page.
|
The decline of inflectional endings on Old English to-infinitives
|
That's the end of this story. The dinosaurs died out; the telegraph died out; the dative ending
on English infinitives died out.
And - just like the dinosaur or the telegraph - it won't be coming back!
Rest in peace, Old English inflected infinitive.