Speaking of Words
By Michael Ferber

            It sounds absurd even to ask if all languages are alike, or even somewhat alike, as anyone studying a foreign language for the first time will attest.  Who could say such a thing?

            Well, when he spoke at UNH in 1995, Noam Chomsky, the most influential linguist of our time, said, “If a Martian linguist visited Earth and listened to humans speaking everywhere, it would conclude that humans all speak the same language, with only trivial dialectal differences among them.”  (I am quoting from memory so I might be off by a word or two.)  All human languages, he says, are really just different dialects of the same language.  This sounds outrageous, but it is only a more pungent version of the claim he has made from the outset of his life-long work on “universal grammar”: the features of syntax, sound, and semantics that any natural language must possess if it is to be intelligible at all, and that all languages on the planet do in fact possess.  He has especially focused on syntax, but he has also studied sounds (phonology) and meanings.

            Meanwhile, independently of Chomsky, Joseph Greenberg has published a list of “language universals” that he derived from a study of thirty disparate languages.  Some he claims are absolute, such as “all languages have pronouns”; some absolute ones are implicational, such as “If a language has a dual number (such as Ancient Greek and Sanskrit) then it must have a plural”; and both kinds may be statistically highly prevalent but not absolutely universal, such as “If a language is of the subject-object-verb type (the most frequent type), then “prepositions” follow the noun or noun-phrase (that is, they are really “postpositions”).”  This seems to be true of many SOV languages, but it was not true of Latin, which was verb-final but had prepositions.

            Chomsky mentioned one in his talk.  German has a “verb-second” rule: the finite verb in an independent clause must be the second element (unless it is in a question).  You may say the equivalent of “I must go tomorrow” or “Tomorrow must I go,” but not “Tomorrow I must go.”  This last sentence puts the finite verb (“must”) third.  Other languages have rules like this, but no language has a “verb-third” rule.  “Our language faculty cannot count,” he said.   (He was not referring to professors in the department of languages.)

            It is claimed that all languages have nouns and verbs, and it certainly seems that without them a language could hardly function as a language, for the world consists of things and events, doesn’t it?  Yet the polysynthetic languages, many of them Native American, may raise doubts, since they make sentences by incorporating “bound elements” that do not exist independently, often in long strings, and it is not clear if the noun-like and verb-like elements are really nouns and verbs.

            Critics of the “universal grammar” project have sometimes argued that the examples that Chomsky and others have offered are too narrowly confined to Indo-European or Semitic languages, while a survey of more remote language families would pour cold water on many proposals.  If there are some 7000 languages in the world (and the number will vary a great deal depending on how strictly you define language and dialect), it is hard to be confident that any candidate for an absolute universal feature won’t be defeated by a single stubbornly strange language.

            Chomsky is associated with the claim that a certain kind of “recursion” can be found in all languages.  A sentence (or complete clause) can recur within a larger sentence, as in “I said I missed her.”  In theory such recursion can go on indefinitely—“He said you thought I believed she loved him”—but of course there are practical limits.  You would think recursion of this sort must be a baked-in feature of any natural language, but Daniel Everett has insisted that the Piraha language, spoken by some 300 people in the Amazon region of Brazil, lacks this capacity.  Other linguists have argued that Everett is mistaken, but so few linguists have studied Piraha that the question is still unsettled.

            It is safe to assert that all spoken languages are based on phonemes, which are the minimal meaningful sounds with which a language makes an indefinitely large number of syllables and words.  In English there are 44 phonemes, including the diphthongs or double vowel sounds.  It is especially rich in fricatives—eight of them—or nine if you count h.  Piraha may have as few as nine phonemes, though this too is still under debate.  The Polynesian languages, such as Hawaiian, have small sets of phonemes; Hawaiian has only eight consonants.  At the other extreme, the !Kung language of Africa has 141 phonemes, including a great number of clicks.  The number of phonemes in a language is normally much less than the possible sounds a speaker might make; we can make several click sounds in English, for example, such as the sucking sound made by the tongue against the hard palate and written “tsk,” but they are not phonemes in English (though they probably are in !Kung). 

            So all languages build their vocabularies out of phonemes, sometimes a very small set of them.  So what?  It is hard to imagine a language that makes words in any other way, but in theory each word could have a unique sound, resembling a bird call, or an animal howl or growl, where factors such as volume, pitch, timbre, length, tune, repetitiveness, or pattern of variation convey different meanings.  As far as we can tell, however, animal languages, if we can call them languages, have small vocabularies.  Humans alone have somehow hit upon the phoneme, so we can make up as many words as we need to convey any meaning we want.  It is perhaps the greatest human discovery.

            The phoneme and the other likely linguistic universals, however, are very general features.  They still allow for a dizzying variety of languages, or so it seems to us earthlings.  In how much detail we might someday describe universal grammar, and thus how interesting it might be, are still far from settled, and there are skeptics who doubt the whole enterprise.  There is a lot at stake, for if there is such a thing, it may well be built into the human brain, as Chomsky thinks, and that would be very interesting.  We may be born knowing universal grammar, and that may be why we are so quick to learn the local language we hear as babies.  I hope linguists will eventually confirm there is a rich universal grammar, but we are a long way from knowing if it really exists beyond a few large generalities.

            I am happy to hear from readers with questions or comments: mferber@unh.edu.

Michael Ferber moved to New Hampshire in 1987 to join the English Department at UNH, from which he is now retired. Before that he earned his BA in Ancient Greek at Swarthmore College and his doctorate in English at Harvard, taught at Yale, and served on the staff of the Coalition for a New Foreign Policy in Washington, DC. In 1968 he stood trial in Federal Court in Boston for conspiracy to violate the draft law, with the pediatrician Benjamin Spock and three other men. He has published many books and articles on literature, and has a deep interest in linguistics. He is married to Susan Arnold; they have a daughter in San Francisco.