The Language Myth by Vyvyan Evans

Subtitle: Why Language is not an Instinct

(Book Review)

Started: October 13, 2017
Finished: March 6, 2018

Why I Read This Book
A new teacher started work at our school back in October.  He was a linguistics major, with plans to pursue a doctorate in cognitive linguistics (after a gap-year teaching ESL.)

I was impressed.  (You'd be surprised how rare genuine linguistics majors are in the field of ESL.  Most ESL Teachers tend to be like me--history or literature majors who couldn't figure out what else to do with their degrees, and so fell into language teaching).

I tried to start up a conversation with him.  "I don't know as much about it as you do, I'm sure," I said tentatively.  "But I did read The Language Instinct by Steven Pinker, so I know the general idea of Universal Grammar.  Is that the kind of thing you study?"

"No, it's the exact opposite," he answered.  "Steven Pinker and Chomsky are all wrong.  Universal Grammar has been disproved."

He obviously knew more about it than me, and yet this wasn't the impression that I had.  "Really?" I said.  "I thought Universal Grammar was still the dominant theory in linguistics."

"No, the field is totally moving away from that now," he said.  "I'll bring you a book tomorrow."

I indicated some hesitation about committing myself to yet another book.  (My - reading - list - has - gotten - a - bit - out - of control lately).
But when he brought it into work, I flipped through it, and it did actually look very light and readable.  "Yeah," I thought, "Actually I could read this easily enough.  Add it to my list."

...and then 6 months later, here I am with my review (*1).

The Background
The full title of this book is "The Language Myth: Why Language is Not an Instinct".
If your first response is: "Wait, what?  Whoever said language was an instinct?  What is this even talking about?"....then you probably need to go back and start out with The Language Instinct by Steven Pinker.
This entire book is a response to The Language Instinct.  And, while The Language Myth can probably be read on its own (Vyvyan Evans gives enough context to make his arguments stand independently), it works best if you know what Evans is responding to.

The Language Instinct by Steven Pinker was published in 1994.
It's primary purpose was to make Chomsky's theories of Universal Grammar understandable to a popular audience.
Chomsky's theories completely upended the field of linguistics in the 1950s and 60s.  And have continued to dominate much of the research and discussion in the field since then.  But since Chomsky and his colleagues write in such a technical academic style, it was impossible for anyone outside of academia to understand what was going on.(*2)  Until Steven Pinker translated it into common English.

Chomsky and Pinker believe the following:
Children don't actually "learn" the grammar of their native language.  This would be impossible.  The grammar is too difficult to be learnt, and besides, children aren't actually exposed to enough input to account for all the complex structures they acquire.  Therefore children must be born with the grammar already pre-programmed in to their brain.  Regardless of which specific language the children end-up learning, they already start with the basics of grammar pre-programmed in.
Of course grammar is different from language to language.  But if you take a cosmic view of the world's languages, the similarities are more striking than the differences.  All languages have the same basic types of word classes (nouns, verbs, adjectives, etc).  All languages have noun phrases and verb phrases.  And all children intuitively understand which how to organize noun phrases and verb phrases, and how to manipulate noun phrases and verb phrases in sentence transformations (e.g. changing a statement into a question.)
This belief that all languages share a basic grammar, and that this basic grammar is something a child is born with, rather than learned, is called Universal Grammar.

...and if that doesn't entirely make sense, that's because it's hard to summarize a 430 page book in 2 paragraphs.  For the complete explanation, you really need to read the book.  You'll just have to trust me on this, it makes sense when Steven Pinker explains it.

The Language Instinct isn't simply just Pinker repeating Chomsky, however.  Pinker adds in plenty of other observations and data to supplement Chomsky.
Pinker also occasionally disagrees with Chomsky.  The biggest point of disagreement is on evolution.  Chomsky is reluctant to give Darwinian explanations for Universal Grammar, but Pinker is adamant that Universal Grammar can, and should, be explained through evolution.

 I read and reviewed The Language Instinct by Steven Pinker back in 2010. I enjoyed it, and I've been greatly influenced by it since then. (In 2016, I put it on my list of top ten non-fiction books.)
Pinker claimed he was representing the consensus in the field of linguistics, and for the most part, I believed him.
It also helped that many of the other books I've read on linguistics echoed Pinker's assertions.  (For example: "An Introduction to Language" by Victoria Fromkin.)

Nevertheless, I have had some hints through the years that there was some controversy about this book.
In 2014, Angry Soba gave me some reading recommendations that contradicted Pinker. including the article Language is Not an Instinct  by Michael Tomasello [PDF HERE], which I never read thoroughly, but I did have a quick look over, and saw a couple of Tomasello's criticisms. (*3)

Also a few of the books on language learning that I've read recently (such as: How Languages are Learned) have placed the Universal Grammar theory as one of many competing theories for language acquisition, indicating that it is NOT the only dominant theory in the field.

And of course The Kingdom of Speech by Tom Wolfe, which I read this time last year, was highly critical of Chomsky's theories.

So it wasn't a completely out-of-the-blue shocker when I got handed The Language Myth.
Nonetheless, for the past 8 years, ever since having read The Language Instinct, I had considered myself a convert to Universal Grammar.  So accepting that Chomsky and Pinker might have gotten it wrong after all was going to be challenging for me.

Evans versus Pinker
Part of what made The Language Instinct so convincing for me was that Pinker kept reassuring the reader that he was simply conveying the scholarly consensus.  He seemed to be constantly saying, "Don't just take my word for it.  This is what ALL the smart people believe."

Vyvyan Evans is using the exact same tactic on the other side.  In his first few chapters, Evans repeatedly emphasizes how many or most of the world's prominent linguists no longer believe in Chomsky's Universal Grammar.
...On the other hand, Evans also says that he was moved to write this book because too many of the university courses and the dominant textbooks still teach Universal Grammar.
So it's a little bit confusing trying to figure out if Universal Grammar is still in vogue or not.  But the general impression that Evans gives is that the field of linguistics is moving away from it.

Despite the fact that I had been a Steven Pinker convert, I have to admit that there were always a few leaps of faith in the theory of Universal Grammar.  I had always known that the search for language universals had been disappointing, and that the evolutionary theory behind universal grammar was problematic.
Nevertheless, when I was reading Steven Pinker, I was so caught up in his authoritative voice that I pushed these doubts to the side of my mind.  (He's a very eloquent speaker and writer.  Just listen to him talk about language and evolution HERE).
A lot of our beliefs are probably like this.  If we're caught up in the eloquence of the speaker, we push the problems to the side of our brain.
Until, that is, someone arrives to point out the problems in a loud and clear voice.  And then what we had avoided thinking about before seems so obvious to us in retrospect.

Evans is such an author for me, and he forced me to take a look at what I thought I knew about language acquisition.  And I think he scores a number of good points against Pinker.
There were other areas, however, where I thought Evans did not offer adequate explanations.
I'll try to outline below where I thought Evans was convincing, and where I thought he was less convincing.
Now, keep in mind throughout this review that I was a history major at school.  I have no idea what I'm talking about when it comes to linguistic controversies.  I'm just your average reader with some opinions. So take my opinion with a huge grain of salt.  Or let me know in the comments where I'm going wrong.

Where Evans Is More Convincing
* Evans makes the point that the theory of Universal Grammar was never based on hard data.  It was simply assumed that languages must have universals, but the theory was never based on compiling data and studying universal grammar patterns.  And in fact, Evans argues, now that the data has come in 50 years later, and we have the evidence of cross-linguistic studies, all available evidence shows that there are no grammatical universals that are common across all languages.

* Not only was Universal Grammar not based on any data about cross-linguistic studies, it was never based on any good neuroscience data.  It made assumptions about a language module in the brain, and grammar that was hard wired in the brain, but never discovered the actual module or the brain circuits.  And, Evans claims, our current understanding of neuroscience shows that the genes can't hard wire information as complex as Universal Grammar into the brain.

* Evolution--Universal Grammar just doesn't make sense within an evolutionary framework.  How could something as complex as grammar evolve?  And moreover, how could grammar have evolved as a beneficial adaptation in one organism, since you need two people in order to communicate?
...although I'm going to moderate this praise with a complaint. Evans is playing a little bit dirty here, since he's deliberately going after the softer targets.  Chomsky has a very unsophisticated view of how evolution could work with Universal Grammar.  Steven Pinker, on the other hand, took great pains to differentiate his view on evolution from Chomsky, and present a more detailed view of how Universal Grammar could have evolved.
So what does Evans do? Ignore Pinker's explanation of evolution entirely, and just attack Chomsky.
Evans repeatedly ridicules Chomsky's idea of a "grammar gene", despite the fact that Pinker made perfectly clear that in his view, grammar was not controlled by one gene. (*4)
Since Evans has obviously read Pinker's book, it's a little bit disingenuous of him to ignore Pinker's more nuanced account of evolution.
Nonetheless... that criticism aside, I think Evans has still done a convincing job of showing that it's very hard to account for Universal Grammar in an evolutionary model.

Where Pinker Is Convincing
The Critical Period Hypothesis:
Evans never says anything about the Critical Period Hypothesis.  He just completely ignores it.
This is a huge chink in his armor.  The critical period is one of the biggest reasons to assume that grammar must have a biological basis.
The Critical Period Hypothesis is still in debate.  But there is a lot of evidence (both anecdotal and in official studies) which indicates that young children learn languages effortlessly and fluently, whereas adults learn languages only through a lot of hard effort, and never reach true fluency.
One possible explanation for this is that the brain's language uploading software is only available for a limited time.  Pinker explains in The Language Instinct that it is inefficient for our brains to carry around the language uploading software for our whole lives, when it is only necessary in early childhood.  So there is a window of time during which children can learn grammar effortlessly, simply through exposure.  And then the brain rewires, and those neurons are used for other purposes. (*5)

Evans even mentions the critical period when he talks about feral children.
In history there have been a handful of cases of feral children--children who have been discovered in the wild around the age of 10 (raised by wolves, etc), who never had any language input during their critical period.  These children never learned to fluently speak a language.
Both Evans and Pinker talk about feral children.  Pinker uses feral children to prove his point that grammar must be biological, because it is connected to a biological critical period.
Evans uses feral children to prove his point that language is socialized, and not purely instinctive.

But Evans completely drops the ball on this, because he doesn't even address the seemingly biological component of the critical period.
I don't think Evans is stupid--he clearly knows what he's talking about.  This omission must have just been an oversight on his part.  But it's still a big omission.

Creating Grammar:
Pinker also makes the case that in newly formed languages, children instinctively create grammatical complexity where none existed before.
Pinker uses the examples of how pidgins become creoles.  And a particular example of a sign language school in Nicaragua, in which the older children created a rudimentary language to communicate, but the younger children instinctively added grammatical complexity.
Evans never really counters this.
Although, interestingly enough, Evans does mention the case of Nicaraguan sign language example, but to illustrate a different point.
As with the feral children mentioned above, it is interesting how Pinker and Evans can both use the same examples to prove completely different points.  But in Evans's case, he has the advantage of writing second. (He is responding to Pinker's book, after all, not vice-versa).  So Evans really should have explained why he thought Pinker was mis-interpreting these case-studies.  Unfortunately, Evans repeatedly misses these opportunities.

The Complexity of Grammar:
The main appeal of the theory of Universal Grammar is that it explains how young, cognitively immature children can perform complex grammatical operations effortlessly.
Chomsky's personal favorite example is "The man who is tall is happy."  In order to change the statement into a question, you need to invert the auxiliary verb (is) with the subject.  But in this sentence, there are two auxiliary verbs (The man who is tall is happy.)  Native speakers, even young children, intuitively understand which auxiliary verb is part of the noun phrase, and which auxiliary verb is part of the verb phrase, so they move the second "is" to the front of the sentence, and not the first one.  "Is the man who is tall happy?"

In my own work in ESL, I've discovered that my students have to learn a lot of complex grammatical transformations which I (as a native-speaker) do effortlessly and unconsciously.  My favorite example is relative pronouns.
Two sentences talking about the same thing can be combined with a relative pronoun.  For example, if I want to say:
"This is the cat" and
"The cat ate the rat", I simply substitute "that" for "the cat" and write:
"This is the cat that ate the rat."
This is difficult for most ESL students, who usually want to say "This is the cat ate the rat", and I have to explicitly teach them that when they add another verb onto the object, they need to use a relative pronoun.
But, if I have two sentences like:
"This is the house" and
"Jack built the house", I can rewrite it as:
"This is the house that Jack built."
Now, notice in this case, "the house" is the object of the second sentence, so I have to first replace it with "that" and then move it to the front of the relative clause.  Again, this is difficult for my students, who want to say: "This is the house that Jack built it."  And I have to teach them that there is no "it" because we already changed the object to "that" and moved it to the front of the clause.

Complicated enough yet?
Then, sometimes the textbook wants to teach the students about when they can omit the relative pronoun.  You can omit the relative pronoun if the relative pronoun is representing the object of the clause, and if it is a defining relative clause.
So, in the sentence "This is the man that I love", I can omit "that" because it represents the object position (I love the man) and because the phrase defines who the man is.  So I can just as easily say; "This is the man I love."
But, I can't omit "that" if it is serving as the subject of the clause.  So I can't say: "This is the man loves me" because here "the man" is the subject of "loves me".  So I must say "This is the man that loves me."
I also can't omit "that" if the clause is non-defining.  So I can say "This is the man I love", but not "This is Doug I love."  Because in this case, we already know who the man is.  (His name is Doug.)  So the relative clause isn't helping to define who he is anymore.  So now we have to say "This is Doug, who I love."

When I teach all this to my students, I have to make a big step-by-step worksheet to help them analyze all the factors in deciding whether or not they can omit the relative pronoun.
But you and I, as native speakers, do this all subconsciously.  In an instant.  Without expending any cognitive effort, you know instantly that "This is the man I love" is correct, and "This is the man loves me" is incorrect.
In fact you can't think about it, because you're busy thinking about the meaning of what you want to say, and you don't have any cognitive resources left to attend to the grammar.
So, the grammar of a native speaker magically takes care of itself.

And this is just one of many examples of complex grammatical transformations that you do thousands of times a day without thinking about it.  (The Ling Space has a whole series of videos explaining the complex grammatical transformations behind everyday utterances).

And that's the other thing.  All native speakers use grammar subconsciously.
My sister, who did study linguistics in university, told me there's a saying in language learning: "You swim in your second language, you breathe in your first."
We don't think about sentence construction in our native language any more than we have to consciously regulate our breathing processes.

Vyvyan Evans never produces any sort of model of language learning that accounts for all these complex subconscious grammar rules.
I'm not doubting that an alternative theory to Universal Grammar probably exists.  But Evans didn't explain it.  With the little he says about language learning, it's impossible to account for the complex grammatical transformations that we do thousands of times day.
He talks about children learning to recognize grammatical patterns in the language, but recognizing patterns is one thing, sentence transformation is another.

So, to me, Evans did not convince me that grammar could be learned through normal cognitive processes and pattern recognition.  There's got to be something more complex going on here.

But... he did at least throw enough cold water on Universal Grammar for me to doubt that theory as well.
I mean, if we say that complex grammar patterns like relative clauses and omitting the relative pronoun are pre-programmed into the brain, then we would have to surmise that evolution has supplied us with some sort of gene that regulates when we can omit the relative pronoun.  And that would be ridiculous, wouldn't it?

Intelligence and Language Learning
Evans successfully (in my opinion) attacks some of Pinker's assertions that language is a distinct module in the mind.  But he never addresses one of Pinker's main claims--that general intelligence does not seem to affect first language acquisition.  All children learn their native language (with all its grammatical complexity) whether they're straight A students or D- students.

Human and Animal Communication:
Pinker and Chomsky both believe that the complexity of human grammar is qualitatively different than any communication devices in the animal kingdom. 
Based on everything I wrote above, I'm inclined to agree with them.
Evans has a long chapter (over 30 pages) in which he tries to convince the reader that there are a lot of parallels between human grammar and animal communication.  But I was totally unconvinced by it.  He was really stretching his examples, and nowhere did he demonstrate that animals have anything close to human grammar.

Lots of Other Stuff
There's actually a lot of stuff in this book I never even got around to discussing--whole chapters on Mentalese and linguistic relativism.  But this review is long enough as it is, so I'm going to call it good here.

Connections with Other Books I've Read
So, The Language Instinct.  Obviously.
Vyvyan Evans also goes after some of Pinker's other books, like The Stuff of Thought.
I never reviewed The Stuff of Thought on this blog because I never finished it.  But I did half-finish it.   In my review of Tootsie, Angry Soba pointed out to me in the comments that I was getting the two books confused, and attributing to  The Language Instinct information that was actually in The Stuff of Thought.  
An Introduction to Language by Victoria Fromkin and colleagues is mentioned by Vyvyan Evans as "the world's best-selling textbook on language (Evans p.20), even though Evans is not a fan of this book, because it supports Chomsky's theory of Universal Grammar.
Daniel Everett, the anti-Chomsky intellectual hero of Tom Wolfe's book The Kingdom of Speech, is cited repeatedly by Vyvyan Evans.
David Crystal is also mentioned occasionally.
Moving on to movies I've seen....the case of Nim Chimpsky is discussed in chapter 2.  Vyvyan Evans implies that the reason why Nim Chimpsky never learned human-like grammar may have had to do more with the mishandling of the experiment than with the chimp's innate abilities.  On this point, at least, he may be half right.  The documentary, Project Nim, does at least show that there was a lot of chaos and disorganization behind this experiment.

Footnotes (docs, pub)
(*1) I guess the obvious question is: if the book is so light and readable, how come it took me 6 months to get through 258 pages?
Well, it's not all the book's fault.  Some of it is just me getting distracted by a lot of other stuff, and spending too much time on Youtube.
But perhaps the book is not quite so readable as it initially looked on my first flip-through.
The prose is actually readable enough (Vyvyan Evans writes very readable sentences).  But he does kind of over-explain some of his points... giving too many examples, or too much explanation, which does kind of kill the forward momentum of this book, and increases the temptation to set the book aside and watch Youtube instead.  And I think that's another reason it took me 6 months to finish.
Really, though, it's more my fault than anything. I procrastinated a lot on this book, and  also got distracted by other projects.  The book is  still readable enough to warrant a recommendation.

(*2) I include myself in this category.  In the past, I've tried to read some of Chomsky's academic writings on linguistics, and couldn't make heads-or-tails of it.  And even a professional linguist like Vyvyan Evans agrees with me.  On page 140 of The Language Myth, Evans writes  "...Chomsky is sometimes difficult to understand, even for many professional linguists..."

(*3) What stuck in my mind most from Tomasello's review was his argument that even if you agreed with the model of language acquisition that Steven Pinker was describing, it would still not qualify as an "instinct" because the definition of an instinct is something that the animal would do with or without any outside input.  (Or, as Tomasello puts it: "would appear in ontogeny even if an individual  were raised in isolation from its species-typical set of experiences ") Language, even under the Universal Grammar model, still needs input.  So language cannot be an instinct.
I thought this was a fair enough criticism, but not something devastating to Pinker's argument.  It just means that he got his nomenclature wrong.  It didn't invalidate his whole theory.

(*4) Pinker made this very clear.  In cases where a genetic abnormality was affecting speech production, Pinker said that it appeared that a single defective gene could disrupt grammar, but that didn't mean that grammar was controlled by just one gene alone.

(*5) This is consistent with what appear to be critical periods for other biological functions.  In graduate school I was told about experiments in which scientists sewed shut the eyes of newborn kittens for the first two months of life.  Even after the eyes are opened, the kittens never were able to see, presumably because the critical period had passed, and the brain had rewired those neurons for other things.

Video Review:
Video review HERE and embedded below:



Link of the Day
Noam Chomsky on Trump and the decline of the American Superpower