This preview shows page 1. Sign up to view the full content.
Unformatted text preview: 11
Language Origins, and Development
1 The linguistic what’s and neurophysiological how’s are incomplete without the evolutionary why’s.
Most species are not intelligent, which suggests that natural selection didn’t favor high intelligence. 2 The heart of Chomsky’s argument is that the human brains are predisposed to use certain types of syntax and not others. The language specializations of the brain are not exclusive; the same areas of the brain have a lot to do with inventing oralfacial and handarm movement sequences, and with judging sounds sequence—and these functions probably evolved together.
3 Part of the language instinct could turn out to be something very simple. E.g.: a real fascination of the young human with discovering any hidden patterns in the sensory environment some of which can be remembered by patterns in the brain. 4 Selection favoring one function may well benefit another function. E.g.: Selection for language abilities benefited musical abilities, because it is hard to figure out what evolutionary circumstances would have rewarded fourpart harmony … The minor product may turn into the major one in the long run.
5 Among the first tasks of early childhood are the discovery of four levels of organization in the apparent chaos of the surrounding environment. Infants discover phonemes and create standard categories for them. With a set of basic speech sounds, babies start discovering patterns amid string of phonemes, averaging nine new words every day.
6 Between 18 and 36 months infants start discovering patterns of words called phrases and clauses, adding –
s for plurals, adding –ed for past tense. After syntax, they go on to discover the rules about narrative having a beginning, middle and end. Thus in four years, children “pyramid” four levels organization, each with its own rules that are causally decoupled from the underlying level’s rule.
The word “chair” is not a chair, but the word “word” is a word.
The correct identification of things in the world—
correct in terms of the consequences we predict from them, rather than in any sense of absolute truth—is adaptive, in the evolutionary sense of the term. 8 Misidentifying genes would not make it as far in the future. Because of their evolutionary value, these processes of identification, these fine discriminations in terms of stored sensory impressions, began very early in evolution, long before mammals walked the earth. In us these processes may seem to have reached a higher pitch of refinement.
9 Words represent something somehow. They serve focus our minds on some aspect of reality.
Words have properties. (adjectives, nouns, verbs, …)
My mental representation of ‘apple’ is only a collection of neurons, all of which are also used for other purposes on occasion. Still, they form an organization that functions pretty well for recognizing apples, eating apples, pronouncing “apple” and so forth.
10 Most of the animal calls are analogous to our exclamations.
One of the evolutionary puzzles is how our ancestors made the transition from a few dozen vocalizations, each with an assigned meaning, to our present system of meaningless phonemes (about 40 in English) that have meaning only in combination with each other. 11 We may want to think of standard exclamations—and most primate calls—as involving an older, more primitive system, located far away from those left lateral brain areas that seem to be important in our kind of syntactic language. We may think in terms of a second language system, operating in parallel with an older one, and not necessary an intensification of the first system. The second system could have its origins in something like face recognition and social relationship, rather than producing vocalization.
12 Words must be able to combine one another, at least in the minimal subject/predicate mode (NP/VP). The subject focus on the object of discourse while the predicate on a property/action … We cannot do it with calls. The latter simply trigger readiness for a certain behavior. No way that calls can be linked to one another.
13 It is words, not sentences that dramatically distinguish our species from others: no animal can learn things that fall outside its biological capacity. A word is the combination of a mental representation of something, which may or my not exist in the real world, with a mental representation of a set of symbols (phonetic, orthographic, manual). What one utters are not words, only the orthographic representation of words.
14 Categories of words, such as proper names, are easy for us, but that’s because our brains have some specializations for them in the front end of the temporal lobe, just in front of where the specialization for facial recognition are located. In all mammals the frontal lobe is used to move and prepare for movements, so it isn’t surprising to find verbs there, at least verbs for when one is the actor.
15 Terms like “subject” and “object” can only be defined over a syntax that already exists. Before syntax existed, they were meaningless. For that matter ancestral humans are very unlikely to have had words like “on” or “the”. Even today the first words of children don’t contain them. (e.g. they don’t contain determiners like “some’, ‘few’, “all”, “many”…).
Before syntax all that existed was a kind of protolanguage. Protolanguage looks like the production of apes that have been taught to use sign or symbols, or at early stage pidgin languages. 17 Protolanguage varieties:
Can only string together a small handful of words at a time.
Can leave out any words they feel like leaving out.
Often depart from the customary word order unpredictably and for no obvious reason.
18 Cannot form any complex structure, whether these be complex noun phrases or sentences more than a clause long.
Contain, if they have any at all, only a tiny fraction of the inflections and the “grammatical words”—things such as article, prepositions and the like—that make up 50% of true language utterances. 19 Protolanguage characteristically consists almost exclusively of nouns and verbs, without any modifiers
If adverbs appear they are usually wholeutterance modifiers, not modifiers of single words. This means that all units are of equal value: every words for itself. Hence, no syntactic structure, no parsing.
20 No structure vs. structure John kissed Mary Mary John kissed 21 This shows that’s happening in the brain. If the brain is working in a protolanguage mode each words is sent separately to the part of the brain that controls the motor organs of speech, and each word is uttered separately. There is also a difference in speed between protolanguage users and language ones.
E.g.: In Hawaii the oldtimers spoke about tree times slower than their own kids.
22 If the brain is working in a languagemode, words are put together in whole phrases and clauses and even sentences before they’re sent to the speech organs to be pronounced. The second diagram also illustrates the order in which words are put together. One parses sentences to find out their meaning (cf. resolution of syntactic ambiguity).
23 Parsing is something we all do every time anything is uttered. But it works quite differently depending on whether what’s uttered is language or protolanguage. If it is protolanguage, it is a good question whether one can be said to parse at all. We can’t decide what the structure is if there is any structure. 24 In protolanguage what one does is just the second part of the operation, i.e. determine the meaning directly from the words (and the context). This is much harder since there is no structure to help. With structure we rarely need context to figure out the meaning of truelanguage utterances, whereas we almost always need context to get the meaning of a protolanguage utterance.
25 Empty categories.
In truelanguage the antecedent is always there somewhere in the sentence, and there are rule to help you find it (e.g.: anaphoric resolution, ellipsis). In protolanguage the empty categories in nowhere in the utterance to be found. Without structure, there cannot be empty categories. 26 If all there was an “epigenetic rule” that said “Seek structure amid chaos”, there would be no Creole languages. Creole languages come into existence when parents who speak a structureless earlystage pidgin pass it on to their children.
The children change the original pidgin in a single generation into a fullfledged language. If they were seeking structure in the pidgin, they wouldn’t find any— they impose structure from within their own minds. 27 We may have acquired the capacity to create structure in language and that capacity then generalized to apply in other spheres.
The language function was likely to be mixed up, locationwise, with some other functions—that “language cortex” isn’t only doing language tasks. There is an enormous overlap with oralfacial and handarm sequencing, for example, suggesting that improvements in one might have benefited the others, at least at some stage in hominid evolution.
28 Phrases and clauses
They were born as twins, and something different has to underlie both phrases and clauses. Argument structure does it: the basic task of language is telling you who did what to whom (as well how and occasionally why). Each participant in the state of action has a specific role to play (PATIENTS, TEHMES and GOALS, i.e. thematic roles). 29 Before there was syntax there were only semantics. So if one looks for the first stages in the development of syntax one must look in semantics for whatever is the most syntaxlike thing. Argument structure is the most plausible candidate (e.g.: PATIENT vs. AGENT). 30 Aphasic patients
Some can swear like sailors. So exclamation seems to survive when the usual lateral language areas (just above, and in front of, your left ear) are damaged. Only damages to a rather distant area of cortex, in the midline above the corpus callosum, affects exclamations as well—and that’s about where the cortical aspect of monkey vocalizations seem to live, too.
31 This seems to suggest that our kind of language didn’t arise from some intensification of the usual ape vocal repertoire.
Several reasons for that: on top of the distance between the two areas, there is also the meaninglessness of our phonemes, in contrast to onesoundonemeaning obligatory interpretation of chimpanzees’ vocalization.
32 The emergence of Protolanguage
It was likely to be a language system similar to what can be taught to apes.
Homo habilis and Homo erectus were probably closer to apes in their behavior … They had brains bigger than those of apes, brain that grew steadily throughout the last several millions years to a size within the range of modern humans, made tools etc. None of this makes it certain that they had protolanguage then, but, given that they brains were not that much smaller than ours, it is reasonable to suppose that they did.
33 There must have been an intermediate state between no language and full language.
Any theory of cooperation suggests that, as a prerequisite, you have to be able to identify individuals. Unspoken proper names—and particularly for individuals that you don’t see everyday—might be a good start for evolving words. 34 Proper names were also needed to keep track of those individuals for mutual support and judge social interactions/situations. Cf. “signature whistle” of dolphins that seem to be used like names for identification purposes. But getting names doesn’t buy you words: monkeys already recognize one another as individual and indeed have a clear map in their head that tell them which individuals are related to one another, without any kind of language to help them.
35 Social intelligence
It’s the favored explanation for the trigger that set language going.
The most remarkable thing about apes society is that it does so closely resemble our own. We see the same maneuvering for status, the same family feuds, we see parental and filial affection, the forming and reforming of alliances, altruism, loyalty, revenge, betrayal.
36 Acts of calculation and deception on the part of several primate species become more frequent among species closer to us.
It seems hat the act of staying successfully selfish within a social community, and competing with individuals some of whom were as smart or smarter than oneself, required a lot more intelligence than hunting or making tools. 37 Complex social interactions/ competitions/ deceptions… may have triggered a theory of mind.
This is a prerequisite for language as we know it.
The capacity to understand another as an intentional individual seems to be linked to linguistic capacity. 38 From protolanguage to language
Once protolanguage had emerged and has reached an appropriate (not necessarily a very high) level of sophistication, it was enthusiastically coopted for manipulation, deception, enhancement of individual prestige, social grooming, gossip, and all the other functions that social intelligence theorists have rightly assigned to it. 39 There are, though, good reasons to think that protolanguage has very little to do with social intelligence.
If you take language away, it’s far from clear that our social lives are any more complex than those of chimpanzees or bonobos.
The selective pressure for language has to come from something that was unique to hominids and something that required the exchange of factual information.
40 We have to look at hominid ecology and how it
differs from that of apes, ancient or modern. Apes live mostly in heavily forested regions of the wilderness tropics.
This mean that in their daily life they don’t have to devote the time and energy that many creatures do to the business of watching for and evading predators. Chimps aren’t strictly vegetarian, but meat is a rare luxury.
41 On both predation risk and food availability the lives of the first hominid were very different. They inhabited grassy savannas with isolated stands of trees and what “gallery woods”. There is an obvious tradeoff between bipedal walking and the ability to climb trees. Our ancestors start walking on two legs at least 2 millions years before any serious brain enlargement, so they were probably relatively poor tree climbers by the time even protolanguage emerged.
42 Savannas were, like today, prime predator country. Ancestors of 2 millions years ago, however, were much smaller than we are. Yet, just like us, they lacked the natural offensive weapons. A species so illequipped would soon have gone extinct if its members hadn’t devoted far more time to predator detection and predator avoidance than apes do. 43 Pragmatic intelligence
Predator detection and predator avoidance depend not on social intelligence but on what we might call pragmatic intelligence.
This is the noting of interpretation of clues in the environment (footprint, crushed vegetation, and so on), something that apes don’t seem to have.
44 Social life didn’t get more complex for our remote ancestor. What did get more complex was the interaction between our ancestors and their environment. The strongest selective pressure would have come from the brute exigencies of survival. 45 Anything, even the crudest and most limited form of protolanguage consisting perhaps of a handful words and/or gestures, has to have had an immediate payoff for the individuals who used it. If it didn’t the behavior wouldn’t have continued and certainly not have been fixed into the genotype. 46 Indirect evidence/support: Most of the first 50 words any child learns are nouns like “leopard”. Not words like “hello”, “good bye”, “please”, “thank you”, we would predict if the language had arisen to cement social networks. Cf. also hunting human accuracy compared to that of primates.
It’s the distinctive feature of human language. It derives from social intelligence. People usually didn’t sharply divide the emergence of protolanguage from the emergence of syntax. These are entirely different things even if one did eventually lead to the other. The events weren’t even close in time.
48 Syntax and semantics
Syntax couldn’t have emerged as a pure novelty. There was semantic before syntax, and if some aspect of semantics could be expressed in terms of syntax, then these aspects make the prime suspect for the source of syntax. 49 Semantic/thematic roles: AGENT, THEME, GOAL. E.g.: With verbs like “sleep”, “run” you have to express one role
With verbs like “make”, “break” you have to express two roles.
With verbs like “give”, “persuade” you have o express three roles.
50 Given a verb one knows in advance if one ought to look for one, two, or tree obligatory arguments.
The number of thematic roles is in the vicinity of seven. The more often obligatory are: AGENT, THEME, GOAL. Then we have the optional ones: TIME, PLACE, BENEFICIARY, and INSTRUMENT.
51 Syntax began when people began to map thematic roles onto their protolinguistic output.
By mapping argument structure onto utterances, you get recursion for free. And recursion is one of the defining characteristics of true language:
S → NP VP
VP → V (NP)
52 Linearization is not something we wanted: it has been forced to us by the physical medium for linguistic communication: we cannot make more that one sound at a time.
The really crucial relationship in language are not horizontal but vertical: horizontal linear relationships cannot explain why “Bob’s sister hurt herself” is grammatical while “Bob’s sister hurt himself” isn’t.
53 There are very fast links between verbs and the nouns people habitually associate with them: if you say “knife” one says “cut”, “bicycle”“ride”, …
Argument structure on its own cannot remove all ambiguities from syntax. Hence the need of parsing. 54 Minimalism
Chomsky removed the distinction between deep and surface structure. We now have just one level of syntax which is a projection of the lexicon. This mean that in the dictionary of your own language that you carry around in our head, stored in the distributed patterns of neural resonances, there is stored along with each words all the features of that words. 55 A word’s features include its meaning, its number and gender (if it has one), the wordclass or classes it belongs to, its function (if it’s a grammatical morpheme), the thematic roles it assigns (if it’s a verb), etc. Some of these features take the form of requirements (e.g.: “the” requires a NP after it). What happens is that one tries to merge words to make larger units (phrases and clauses) by matching features. If the positive features of one word match the requirements of the other, you can merge them and move onto the next merger. If not, the derivation crashes.
56 This model is a lot closer to how the brain works, i.e. how it handles language than the old (deep/surface structure) model. It makes no neurological sense to say that the brain first shaped up some very abstract sentence structure and then fooled around with them to make something completely different come out of your mouth, which is that the old deepstructure/surfacestructure model was implicitly claiming. Brains aren’t that subtle. If they can do it the straightforward way, they do it the straightforward way.
View Full Document
- Fall '07