Take a photo of a barcode or cover
A review by blurred
Language: The Cultural Tool by Daniel L. Everett
3.0
A decent, readable book on the nature of language that succeeds in exploring a lot of interesting ideas about how language is acquired and used. Unfortunately, though, Everett's argumentation is often unfocussed and discursive, and I often found myself wondering what point he was really trying to make in this book. His main target throughout seems to be "nativism" - the idea that language, or the cognitive capacities that lead to language, is somehow innate - but the target of his invective is largely a strawman, and not representative of any position that I've ever encountered in any linguistic theory.
Take, for example, this quote from page 70:
We must note at the beginning that there is not a single, scientifically informed nativist on the planet who would accede to the claim that there is a "gene for language". No-one expects there to be, and no nativist would posit this as a reasonable expectation. That's simply not how genes work, as Everett kind of points out here. Genes - particularly those involved with the development of the brain, for some reason - tend to be pleiotropic; that is, they are expressed in the development of more than one "phenotype", or bodily feature. As Everett suggests later, it may well be the case that genes identified as being integral to the development of language (such as the FOXP2 example given here) are also integral to other areas of ontogenic development, and cannot therefore be naively classified as "language genes". However, this rather misses the point somewhat.
The first issue here is that no gene is, strictly, a gene for anything. All genes do is produce certain proteins at certain stages of ontogenic development, and whatever the effect this protein has is only discernible as part of a longer, cumulative process. That is, it only finds phenotypical expression through interaction with other genes, which are themselves doing nothing other than producing proteins. On the other hand, there are clearly some genes that are necessary for the emergence of particular phenotypes. Eye colour would be a fairly obvious example here. My eyes are blue because I happen to possess a particular recessive allele that people with brown eyes do not have. In that respect, it wouldn't be wildly incorrect to say that this particular gene is a gene for blue eyes. Strictly speaking this isn't true, of course (this gene doesn't create "blue eyes" it merely supplies a protein, which may also be expressed in phenotypes other than that of eye-colour) but in another, less pedantic sense we can say that this gene is apparently necessary for the development of blue eyes, and - to that extent - it is most certainly a gene for blue eyes. If nativists slip and call genes such as FOXP2 "genes for language", then that is all they really mean: namely, that such genes are necessary for the development of language, regardless of their precise role in the ontogeny of neural development, or what other role they may have in any other kind of phenotypical development.
The second issue concerns why a particular gene is present in our genome. This is a controversial area in evolutionary biology, but generally if we find a gene that is preserved with almost universal fidelity in a particular gene-pool, then we are probably justified in presuming that there have been positive selection pressures in the evolutionary history of the organism, leading to its preservation in the gene pool. (This idea is controversial because neutral selection patterns - such as "genetic drift" - can also have a large effect on the genome, leading some biologists to denounce the search for past positive selection pressures as little more than teleological sophistry. Such arguments are a little beside the point here, though.) If we ask why a gene such as FOXP2 has come to be so universally present in the human genome in such a small space of time, then we are probably justified in asking what positive benefit it accrued to our ancestors that led to its ubiquity. Given the importance of language to our species - and the negative costs that must have been associated with its absence - I think we would probably be correct in asserting that the prevalence of the FOXP2 gene in the human genome is linked primarily to the very strong selection pressures that relate to language use, rather than whatever other functions the gene may serve. In other words, regardless of its precise role or function in the human organism, the rapidity with which it became ubiquitous in the human genome can most probably be attributed to positive selection relating to its importance to the development of language. Again, to that extent, it is a "gene for language".
From such claims, Everett proceeds to denounce more biologically reductive theories of language acquisition, in favour of more plastic, cultural explanations. Again, I found such claims somewhat wanting. Plainly there must be some biological basis for language (as Steven Pinker points out, this is why a child raised in a certain household will acquire language, whereas a dog living in the same household will not) and the fact that there is a large cultural component to language doesn't change this. (Furthermore, I would ask whether or not we might be able to claim that culture itself has a biological basis!) He largely succeeds in showing that some of the more excessive nativist claims concerning grammar are misguided by demonstrating the sheer plurality of different forms that grammar can take, but I don't think that the more restrained claims of nativism - for example, the universality of recursion in language - are necessarily challenged by this. Again, Everett might have been a little more successful in my eyes if he had narrowed the scope of his argument a little, rather than trying to take down the entire edifice of nativist linguistics in one go.
But if we take out the needless and slightly confused swipes at scientific theories of language acquisition, there's little else in the book that strikes me as particularly controversial. Everett runs through the peculiarities of different languages (as a riposte to Chomskyan theories of "universalist" grammar) and theories of language acquisition stretching from Plato to the current day. His main idea, such as I could tell, was that language is a "tool", and like all other tools its specific manifestation among specific groups of people depends largely on the function it is required to fulfil. In other words, language is a highly pliable tool that has not arisen as a consequence of biological determinism, but rather as a consequence of "fitness for purpose" that finds its expression in a number of ways. I didn't find his arguments convincing all the time, but I will concede that he raised a number of interesting points throughout the book, and as such it would still be worth reading even if you don't feel you're likely to agree with what he has to say. It would have been a more successful book, though, had it retained a more clear focus.
Take, for example, this quote from page 70:
It has not been established that there are any genes specific to language. What we do know from genetic studies so far is that there are genes – the best example, widely discussed in the literature, being FOXP2 – that are important to language. Yet finding a gene that is important to language is not the same as identifying a gene for language.
We must note at the beginning that there is not a single, scientifically informed nativist on the planet who would accede to the claim that there is a "gene for language". No-one expects there to be, and no nativist would posit this as a reasonable expectation. That's simply not how genes work, as Everett kind of points out here. Genes - particularly those involved with the development of the brain, for some reason - tend to be pleiotropic; that is, they are expressed in the development of more than one "phenotype", or bodily feature. As Everett suggests later, it may well be the case that genes identified as being integral to the development of language (such as the FOXP2 example given here) are also integral to other areas of ontogenic development, and cannot therefore be naively classified as "language genes". However, this rather misses the point somewhat.
The first issue here is that no gene is, strictly, a gene for anything. All genes do is produce certain proteins at certain stages of ontogenic development, and whatever the effect this protein has is only discernible as part of a longer, cumulative process. That is, it only finds phenotypical expression through interaction with other genes, which are themselves doing nothing other than producing proteins. On the other hand, there are clearly some genes that are necessary for the emergence of particular phenotypes. Eye colour would be a fairly obvious example here. My eyes are blue because I happen to possess a particular recessive allele that people with brown eyes do not have. In that respect, it wouldn't be wildly incorrect to say that this particular gene is a gene for blue eyes. Strictly speaking this isn't true, of course (this gene doesn't create "blue eyes" it merely supplies a protein, which may also be expressed in phenotypes other than that of eye-colour) but in another, less pedantic sense we can say that this gene is apparently necessary for the development of blue eyes, and - to that extent - it is most certainly a gene for blue eyes. If nativists slip and call genes such as FOXP2 "genes for language", then that is all they really mean: namely, that such genes are necessary for the development of language, regardless of their precise role in the ontogeny of neural development, or what other role they may have in any other kind of phenotypical development.
The second issue concerns why a particular gene is present in our genome. This is a controversial area in evolutionary biology, but generally if we find a gene that is preserved with almost universal fidelity in a particular gene-pool, then we are probably justified in presuming that there have been positive selection pressures in the evolutionary history of the organism, leading to its preservation in the gene pool. (This idea is controversial because neutral selection patterns - such as "genetic drift" - can also have a large effect on the genome, leading some biologists to denounce the search for past positive selection pressures as little more than teleological sophistry. Such arguments are a little beside the point here, though.) If we ask why a gene such as FOXP2 has come to be so universally present in the human genome in such a small space of time, then we are probably justified in asking what positive benefit it accrued to our ancestors that led to its ubiquity. Given the importance of language to our species - and the negative costs that must have been associated with its absence - I think we would probably be correct in asserting that the prevalence of the FOXP2 gene in the human genome is linked primarily to the very strong selection pressures that relate to language use, rather than whatever other functions the gene may serve. In other words, regardless of its precise role or function in the human organism, the rapidity with which it became ubiquitous in the human genome can most probably be attributed to positive selection relating to its importance to the development of language. Again, to that extent, it is a "gene for language".
From such claims, Everett proceeds to denounce more biologically reductive theories of language acquisition, in favour of more plastic, cultural explanations. Again, I found such claims somewhat wanting. Plainly there must be some biological basis for language (as Steven Pinker points out, this is why a child raised in a certain household will acquire language, whereas a dog living in the same household will not) and the fact that there is a large cultural component to language doesn't change this. (Furthermore, I would ask whether or not we might be able to claim that culture itself has a biological basis!) He largely succeeds in showing that some of the more excessive nativist claims concerning grammar are misguided by demonstrating the sheer plurality of different forms that grammar can take, but I don't think that the more restrained claims of nativism - for example, the universality of recursion in language - are necessarily challenged by this. Again, Everett might have been a little more successful in my eyes if he had narrowed the scope of his argument a little, rather than trying to take down the entire edifice of nativist linguistics in one go.
But if we take out the needless and slightly confused swipes at scientific theories of language acquisition, there's little else in the book that strikes me as particularly controversial. Everett runs through the peculiarities of different languages (as a riposte to Chomskyan theories of "universalist" grammar) and theories of language acquisition stretching from Plato to the current day. His main idea, such as I could tell, was that language is a "tool", and like all other tools its specific manifestation among specific groups of people depends largely on the function it is required to fulfil. In other words, language is a highly pliable tool that has not arisen as a consequence of biological determinism, but rather as a consequence of "fitness for purpose" that finds its expression in a number of ways. I didn't find his arguments convincing all the time, but I will concede that he raised a number of interesting points throughout the book, and as such it would still be worth reading even if you don't feel you're likely to agree with what he has to say. It would have been a more successful book, though, had it retained a more clear focus.