How did language begin? - Linguistic Society of America

How didlanguagebegin?Written by Ray Jackendoff

What does the question mean?In asking about the origins of human language, wefirst have to make clear what the question is. Thequestion is not how languages gradually developedover time into the languages of the world today.Rather, it is how the human species developed overtime so that we — and not our closest relatives, thechimpanzees and bonobos — became capable ofusing language.And what an amazing development this was! Noother natural communication system is like humanlanguage. Human language can express thoughts onan unlimited number of topics (the weather, the war,the past, the future, mathematics, gossip, fairy tales,how to fix the sink...). It can be used not just toconvey information, but to solicit information (questions)and to give orders. Unlike any other animalcommunication system, it contains an expression fornegation — what is not the case. Every human languagehas a vocabulary of tens of thousands ofwords, built up from several dozen speech sounds.Speakers can build an unlimited number of phrasesand sentences out of words plus a smallish collectionof prefixes and suffixes, and the meanings ofsentences are built from the meanings of the individualwords. What is still more remarkable is thatevery normal child learns the whole system fromhearing others use it.Animal communication systems, in contrast, typicallyhave at most a few dozen distinct calls, and they areused only to communicate immediate issues such asfood, danger, threat, or reconciliation. Many of thesorts of meanings conveyed by chimpanzee communicationhave counterparts in human 'body language'.For animals that use combinations of calls(such as some songbirds and some whales), themeanings of the combinations are not made up ofthe meanings of the parts (though there are manyspecies that have not been studied yet). And theattempts to teach apes some version of human language,while fascinating, have produced only rudimentaryresults. So the properties of human languageare unique in the natural world.How did we get from there to here? All present-daylanguages, including those of hunter-gatherer cul-

tures, have lots of words, can be used to talk aboutanything under the sun, and can express negation.As far back as we have written records of humanlanguage — 5000 years or so — things look basicallythe same. Languages change gradually over time,sometimes due to changes in culture and fashion,sometimes in response to contact with other languages.But the basic architecture and expressivepower of language stays the same.The question, then, is how the properties of humanlanguage got their start. Obviously, it couldn't havebeen a bunch of cavemen sitting around and decidingto make up a language, since in order to do so,they would have had to have a language to startwith! Intuitively, one might speculate that hominids(human ancestors) started by grunting or hooting orcrying out, and 'gradually' this 'somehow' developedinto the sort of language we have today. (Such speculationswere so rampant 150 years ago that in 1866the French Academy banned papers on the origins oflanguage!) The problem is in the 'gradually' and the'somehow'. Chimps grunt and hoot and cry out, too.What happened to humans in the 6 million years orso since the hominid and chimpanzee lines diverged,and when and how did hominid communicationbegin to have the properties of modern language?Of course, many other properties besides languagedifferentiate humans from chimpanzees: lowerextremities suitable for upright walking and running,opposable thumbs, lack of body hair, weaker muscles,smaller teeth — and larger brains. According tocurrent thinking, the changes crucial for languagewere not just in the size of the brain, but in itscharacter: the kinds of tasks it is suited to do — asit were, the 'software' it comes furnished with. Sothe question of the origin of language rests on thedifferences between human and chimpanzee brains,when these differences came into being, and underwhat evolutionary pressures.What are we looking for?The basic difficulty with studying the evolution oflanguage is that the evidence is so sparse. Spokenlanguages don't leave fossils, and fossil skulls onlytell us the overall shape and size of hominid brains,not what the brains could do. About the only defini-

tive evidence we have is the shape of the vocal tract(the mouth, tongue, and throat): Until anatomicallymodern humans, about 100,000 years ago, theshape of hominid vocal tracts didn't permit the modernrange of speech sounds. But that doesn't meanthat language necessarily began then. Earlierhominids could have had a sort of language thatused a more restricted range of consonants andvowels, and the changes in the vocal tract may onlyhave had the effect of making speech faster andmore expressive. Some researchers even proposethat language began as sign language, then (graduallyor suddenly) switched to the vocal modality,leaving modern gesture as a residue.These issues and many others are undergoing livelyinvestigation among linguists, psychologists, andbiologists. One important question is the degree towhich precursors of human language ability arefound in animals. For instance, how similar are apes'systems of thought to ours? Do they include thingsthat hominids would find it useful to express to eachother? There is indeed some consensus that apes'spatial abilities and their ability to negotiate theirsocial world provide foundations on which thehuman system of concepts could be built.A related question is what aspects of language areunique to language and what aspects just draw onother human abilities not shared with other primates.This issue is particularly controversial. Some researchersclaim that everything in language is built out of otherhuman abilities: the ability for vocal imitation, the abilityto memorize vast amounts of information (both neededfor learning words), the desire to communicate, theunderstanding of others' intentions and beliefs, and theability to cooperate. Current research seems to showthat these human abilities are absent or less highlydeveloped in apes. Other researchers acknowledge theimportance of these factors but argue that hominidbrains required additional changes that adapted themspecifically for language.Did it happen all at once or in stages?How did these changes take place? Some researchersclaim that they came in a single leap, creatingthrough one mutation the complete system in the

ain by which humans express complex meaningsthrough combinations of sounds. These people alsotend to claim that there are few aspects of languagethat are not already present in animals.Other researchers suspect that the special propertiesof language evolved in stages, perhaps over somemillions of years, through a succession of hominidlines. In an early stage, sounds would have beenused to name a wide range of objects and actionsin the environment, and individuals would be ableto invent new vocabulary items to talk about newthings. In order to achieve a large vocabulary, animportant advance would have been the ability to'digitize' signals into sequences of discrete speechsounds — consonants and vowels — rather thanunstructured calls. This would require changes in theway the brain controls the vocal tract and possiblyin the way the brain interprets auditory signals(although the latter is again subject to considerabledispute).These two changes alone would yield a communicationsystem of single signals — better than the chimpanzeesystem but far from modern language. A next plausiblestep would be the ability to string together severalsuch 'words' to create a message built out of themeanings of its parts. This is still not as complex asmodern language. It could have a rudimentary 'meTarzan, you Jane' character and still be a lot betterthan single-word utterances. In fact, we do find such'protolanguage' in two-year-old children, in the beginningefforts of adults learning a foreign language, andin so-called 'pidgins', the systems cobbled together byadult speakers of disparate languages when they needto communicate with each other for trade or othersorts of cooperation. This has led some researchers topropose that the system of 'protolanguage' is stillpresent in modern human brains, hidden under themodern system except when the latter is impaired ornot yet developed.A final change or series of changes would add to'protolanguage' a richer structure, encompassing suchgrammatical devices as plural markers, tense markers,relative clauses, and complement clauses (“Joe thinksthat the earth is flat”). Again, some hypothesize thatthis could have been a purely cultural development,

and some think it required genetic changes in thebrains of speakers. The jury is still out.When did this all happen? Again, it's very hard totell. We do know that something important happenedin the human line between 100,000 and50,000 years ago: This is when we start to find culturalartifacts such as art and ritual objects, evidenceof what we would call civilization. What changed inthe species at that point? Did they just get smarter(even if their brains didn't suddenly get larger)? Didthey develop language all of a sudden? Did theybecome smarter because of the intellectual advantagesthat language affords (such as the ability tomaintain an oral history over generations)? If this iswhen they developed language, were they changingfrom no language to modern language, or perhapsfrom 'protolanguage' to modern language? And if thelatter, when did 'protolanguage' emerge? Did ourcousins the Neanderthals speak a protolanguage? Atthe moment, we don't know.One tantalizing source of evidence has emergedrecently. A mutation in a gene called FOXP2 hasbeen shown to lead to deficits in language as wellas in control of the face and mouth. This gene is aslightly altered version of a gene found in apes, andit seems to have achieved its present form between200,000 and 100,000 years ago. It is very temptingtherefore to call FOXP2 a 'language gene', but nearlyeveryone regards this as oversimplified. Are individualsafflicted with this mutation really languageimpairedor do they just have trouble speaking? Ontop of that, despite great advances in neuroscience,we currently know very little about how genes determinethe growth and structure of brains or how thestructure of the brain determines the ability to uselanguage. Nevertheless, if we are ever going to learnmore about how the human language abilityevolved, the most promising evidence will probablycome from the human genome, which preserves somuch of our species' history. The challenge for thefuture will be to decode it.

For further informationChristiansen, Morton H. and Simon Kirby (eds.).2003. Language Evolution. New York: OxfordUniversity Press.Hauser, Marc; Noam Chomsky; and W. TecumsehFitch. 2002. The faculty of language: What is it, whohas it, and how did it evolve? Science 298.1569-79.Hurford, James; Michael Studdert-Kennedy; and ChrisKnight (eds.). 1998. Approaches to the Evolution ofLanguage. Cambridge: Cambridge University Press.Jackendoff, Ray. 1999. Some possible stages in theevolution of the language capacity, Trends inCognitive Sciences 3.272-79.Pinker, Steven, and Ray Jackendoff. 2005. The facultyof language: What's special about it? Cognition95.210-36.

The Linguistic Society of America was founded in1924 for the advancement of the scientific study oflanguage. The Society serves its nearly 6,000 personaland institutional members through scholarly meetings,publications, and special activities designed toadvance the discipline.The Society holds its Annual Meeting in early Januaryeach year and publishes a quarterly journal,LANGUAGE, and the LSA Bulletin. Among its specialeducation activities are the Linguistic Institutes heldevery other summer in odd-numbered years andco-sponsored by a host university.The web site for the Society (http://www.lsadc.org)includes The Field of Linguistics (brief, nontechnicalessays describing the discipline and its subfields)and statements and resolutions issued by theSociety on matters such as language rights, theEnglish-only/English-plus debate, bilingual educationand ebonics.Linguistic Society of America1325 18th St, NW, Suite 211Washington, DC 20036-6501(202) 835-1714lsa@lsadc.orghttp://www.lsadc.orgDuplicate as needed08/06