I have been thinking about last week’s post; it discussed a paper lamenting that the generative approach is ending with a whimper rather than a bang: “insights… recognized since the very origins of generative grammar … seem to have been forgotten, ignored, or even denied without serious argument.” The distinguished authors – Martin Everaert, Mannus Huybregts, Noam Chomsky, Robert Berwick, Johan Bolhuis – appear to have no idea why their movement is no longer at the cutting edge of linguistic activity.
Not surprisingly, they look to changing alignments of the stars rather than wondering how much they themselves might have contributed to their plight. I was especially struck by their complaint that generative theory has been pushed aside “without serious argument.” There has, of course, been tremendous argument, but very little response that has taken the criticism of generative linguistics seriously. Many rival theories have been put forward, their authors eager to dispute the generativists, but they have gotten little to no back and forth.
A typical example is the paper by Nicholas Evans and Stephen Levinson titled “The Myth of Language Universals.” It went after the generativist’s claim that all languages reflect the same hierarchical structure and organizing procedures. In this view, individual languages are an externalized version of what is happening in the brain. All externalized languages must satisfy the basic properties of the internal, universal symbol organizer (called the “universal grammar” or UG in generativist jargon). The Evans-Levinson paper questioned this belief and opened with a direct throwing down of the gauntlet, “Languages are much more diverse in structure than cognitive scientists generally appreciate. A widespread assumption among cognitive scientists, growing out of the generative tradition in linguistics, is that all languages are English-like, but with different sound systems and vocabularies.”* The paper then went on to show in enormous detail that languages are radically varied in syntax. This effort pulled together the long and widespread efforts of linguists to learn and describe the many languages still found around the world. It was a stunning, extensive review of linguistic difference. Nobody expected the generativists to say, “Oops,” and give up on the UG right then and there, but the paper’s appearance did demand a serious effort to show that at least some of these differences could be explained by a universal rule.
This opportunity for rebuttal came at once, for the Evans-Levinson paper appeared in a wonderful journal, Behavioral and Brain Sciences, which includes pro- and con- commentary essays with each original paper it publishes. A number of generativists did supply criticism, but their basic argument was that examples of diversity did not disprove an underlying universality. That point was true, but too simple. Nobody responded to the challenge by showing a universal rule that explained a seemingly diverse language feature.
Let me correct that right away. When I say “nobody,” I mean no generativist. An anti-generativist, Michael Tomasello, did list a number of universals, e.g. “agents of action,” “patients of action,” “possessors” and “locations” but he attributed all of these things to “general cognitive principles” and denied that there were any “biological adaptations with specific linguistic content.” Tomasello was stating exactly the position I have come to support on this blog. Argument got me here.
About a year after I posted a discussion of the original Evams-Levinson paper, I posted a report on a second paper by the same authors that discussed the positive reaction to the first paper. My post drew an anonymous response from a professional generativist: “not all evidence is created equal. Some is more believable than others. So when you have a solid theory supported by a lot of solid data but happens to clash with some fairly weak data, it's not clear that you should keep the weak data and toss out the theory.” This kind of oh-poo response is pretty typical of the way generativists have evaded serious argument for decades. There is in the history of generative linguistics no equivalent to the debate between Louis Pasteur and Fèlix Pouchet over spontaneous generation, nothing in which the generativists tightened and proved their doctrine by refuting a series of pointed challenges. After a while, the suspicion grew among anti-generativists that the other side was ducking a fight because they couldn’t answer the challenges. That suspicion may have been grossly unfair, but it has left the skeptics shaking their heads when they read a paper signed by leading generativists that laments their fading status without ever being offered a fair fight.
My greatest fear is that as the generativist effort falls by the wayside, their accomplishments will be lost as well, and they do have some accomplishments. For instance the generativists have clearly established that phrases are the basic unit of a sentence, and they have identified a series of syntactical peculiarities that must be explained. A speaker, for example, can say I wanna give Jack the ball but not *Who do you wanna give Jack the ball? Why can you pronounce the first want to as a single word but not the second? I have proposed a non-generative explanation, but most anti-generativists have not bothered.
Typically in psychology, schools of thought replace one another without preserving anything from the old ways. In 1959 most scientific psychology was behaviorist psychology. Then Chomsky came along and in one blow ended stimulus-response theories of language. I doubt that it ever crossed any gnerativist’s mind to see if there was anything at all useful to rescue from B.F. Skinner’s examination of language. Will the long generative tradition end on a similar scrap heap?
* If you want to hear Chomsky continue to make this point about the difference in language coming from the externalizing process, see the March 2015 Youtube discussion Chomsky and Krauss 35 minutes into the chat.