Inf1 CogSci 2013: Language 4: Dissection by linguistics
Language is simply the fulfilment of a need to communicate.
For every idea there is a word, and vice versa.
We utter the words in a order that reflects the connection between ideas.
This view is wrong!
Human language is a complex system in itself, with a highly intricate anatomy.
Language does express meaning as sound.
But not as a single step.
There appears to be an "assembly line" of mental modules:
In the previous lecture, we provided evidence for dividing linguistic knowledge into
Furthermore, we suggested that
The word "word" has two distinct senses which we need to keep separate:
The second sense of "word" covers things which appear only as parts of "words" in the first sense:
It might also extend to some super-"word-in-the-first-sense" units, such as idioms, cliches and collocations:
So:
Baudouin de Courtenay's morphemes - "that part of a word which is endowed with psychological autonomy and is for the very same reasons not further divisible".
Morphological rules - build complex words from simple words and morphemes.
Syntactic rules - build phrases and sentences.
Why do linguists regard these two mental modules as distinct?
With some complex nouns, it is unclear what the "correct" plural form is:
The reason that these plural forms are controversial is that both alternatives have some logic on their side.
It depends on how the language user analyses the chunk of sound:
Simple words - add -s to the end
Compounds - again add -s to the end
Phrases - again add -s to the end of the head word
As a language evolves over time, a phrase may get "repackaged" as a word, if it is used particularly frequently:
In these cases we no longer perceive the internal semantic structure of the word.
Controversial plural forms arise when older speakers have not yet repackaged the phrase, but younger speakers have.
Other suffixes get assimilated too:
Some people subconciously analyse "mother-in-law" as including an embedded phrase:
Others treat it as an unanalysable simple word:
Thus, we need to distinguish morphological rules (whose output is a word) from syntactic rules (whose output is a sentence or phrase).
The following plural forms are particularly interesting, since they involve an adjective appearing after the head noun:
These words were all borrowed into Middle English from Norman French (because England was ruled by the Normans from 1066 onwards, and they imposed their political system on the existing population)
Interestingly, we have totally lost the Norman French grammatical logic in the related word "major-general"
English verb inflection is famous for being boring
Spanish and Italian verbs often have as many as 50 distinct forms
The Bantu language Kivunjo has verbs with half a million distinct forms!
The average English verb has just four distinct forms.
But these four forms are shared between at least 13 different syntactic roles.
The null suffix
The -s suffix
The -ing suffix
The -ed suffix
Syncretism = inflectional miserliness
Irregular verbs can have a few more forms.
Many have distinct past tense and past participial forms:
A few have a distinct form for the verbal adjective:
And one English verb has eight distinct forms:
Syncretism - one form, several roles
Allomorphy - one role, several forms
For example, the regular past tense suffix, written -ed, is actually pronounced in three different ways:
The same thing goes for the regular plural:
However, we don't need to say there are three different regular past tense suffixes, and three different regular plural suffixes.
The mental lexicon contains just the one suffix for each role, and the rules of phonology fiddle with its pronunciation.
The lexicon contains:
There is a morphological rule:
And there are two phonological rules which tweak the inflected words to make them easier to pronounce:
Therefore, we don't need to posit three distinct past tense suffixes in English, i.e. [d], [t] and [əd] - these distinctions are captured by generalisations in the phonology module.
In Middle English (1200-1500), the only past tense suffix was [əd]
By Early Modern English, the 'clipped' suffixes [d] and [t] had become commonplace.
Contemporary spelling distinguished these with an apostrophe:
When Samuel Johnson standardised English spelling in the late 1700s, he reinserted the written 'e', even when there is no corresponding vowel in the word's pronunciation:
English inflection is very simple.
Therefore, we can posit just one rule capturing all English verb inflection:
The stems and suffixes themselves are stored in, and retrieved from, the mental lexicon.
All irregular past tense verb forms used to be regular!
However: the process of language acquisition by children has slightly imperfect fidelity:
Hypothesis:For any irregular past tense form in Modern English, some past generation of English speakers must have failed to grasp the relevant rule
Once a past tense form has been stored as a word, it can subsequently pick up more and more idiosyncracies
First assignment is out, due two weeks yesterday
Starting this week, you should review the tutorial handout in advance of your tutorial session
Required reading: Stephen Pinker, Words and Rules: Chapter 2 "Dissection by Linguistics" and Chapter 3, "Broken Telephone" (The first 25 pages or so are online at The publisher's website).