Creole birdsong?

« previous post | next post »

Yesterday, I spent a fascinating afternoon at Ofer Tchernichovski's lab at CCNY. And a couple of weeks ago, the Penn Linguistics Department colloquium featured Ofer talking about some of his lab's recent research, including this work: Olga Feher, Partha P. Mitra, and Ofer Tchernichovski, "Abnormal isolate birdsong evolves into normal song over a few generations".

Zebra finches are among the songbirds who learn their songs by imitating adults, just as human children learn their language by interaction with those who already know it. Male songbirds raised in isolation, without any conspecific adult models during the critical period for song learning, are handicapped for life: they develop only an ill-organized, infantile "subsong". From the example of abused or feral children like Genie, we know that something similar happens with human children.

In both cases, this raises a sort of chicken-and-egg question: if normal development requires an adult model, then which came first, the pupil or the tutor?

One obvious possibility is that the normal pattern is implicit in the species genotype, but requires a combination of cultural evolution and infant learning, repeated over several generations, to develop completely. This concept is discussed in Ann Senghas and Marie Coppola, "Children creating language: How Nicaraguan Sign Language acquired a spatial grammar", Psychological Science, 12(4): 323-328, 2001

It has long been postulated that language is not purely learned, but arises from an interaction between environmental exposure and innate abilities. The innate component becomes more evident in rare situations in which the environment is markedly impoverished. The present study investigated the language production of a generation of deaf Nicaraguans who had not been exposed to a developed language. We examined the changing use of early linguistic structures (specifically, spatial modulations) in a sign language that has emerged since the Nicaraguan group first came together. In under two decades, sequential cohorts of learners systematized the grammar of this new sign language. We examined whether the systematicity being added to the language stems from children or adults; our results indicate that such changes originate in children aged 10 and younger. Thus, sequential cohorts of interacting young children collectively possess the capacity not only to learn, but also to create, language.

The cited work by Olga Feher et al. demonstrates this kind of "multi-generational phenotype" (Ofer's phrase) experimentally, in a colony of zebra finches whose founder was an isolate. As each succeeding generation learns songs from the preceeding one, the effects of biases in the learning process accumulate, so that after a few generations, normal zebra finch songs have re-emerged.

Derek Bickerton's influential theory of how "creole languages" develop — the "bioprogram hypothesis" — was of this type, though it emphasized an abrupt emergence of new patterns in first-language learners, rather than a more gradual development over several generations. From the article on "Pidgins and Creoles" by John Rickford and Barbara Grimes, in OUP's International Encyclopedia of Linguistics:

A pidgin is a lingua franca that arises to facilitate communication between speakers of different languages who are in repeated or sustained contact with each other, as in trade or plantation situations. Although it is no one's native language, a pidgin usually involves mixture or compromise between the native languages of its users; in comparison with these, it is restricted in social role and simplified or reduced in linguistic resources. A creole, by contrast, is a native or first language for its speakers, as in the case of children born to pidgin-speaking adults on a plantation. In the process of being acquired and used as a native language, the linguistic resources of a pidgin became expanded.


As noted above, the linguistic resources of a pidgin or jargon are complicated and extended in the process of creolization—its use as a primary and often a native language. Evidence of creolizing expansion came from Bickerton's 1981 study of the nonnative Hawaiian Pidgin English (HPE) spoken by surviving Filipino, Japanese, and other immigrants who arrived in Hawaii prior to 1920, compared with the Hawaiian Creole English (HCE) spoken natively by their children. In a number of central grammatical areas, HCE contains linguistic resources that HPE does not, and these resources give the former greater referential power and flexibility. The list includes movement rules that allow fronting of sentence constituents for focusing or emphasis (o, daet wan ai si ‘oh, I saw that one’), and the use of preverbal auxiliaries (bin, go, stei) to mark tense, modality, and aspect.

Finding similar features in other creoles—which also seemed to have developed “abruptly” from jargons or pre-pidgin continua—Bickerton hypothesized that creoles represent the workings of a species-specific bioprogram for language that is used by the first creole-speaking generation of children to flesh out their parents' pidgin. This intriguing hypothesis generated considerable discussion, but it no longer commands the attention it once did. In part this is because it does not account for non-European creoles like Lingala and Sango, nor for creoles that emerged from expanded pidgins, like Tok Pisin. Finally, both adults and substratal languages now seem to have played a larger role than provided for in the bioprogram theory, and Haitian and other creoles apparently developed gradually rather than through the abrupt, nativizing process the bioprogram requires (see Rickford and McWhorter 1997).

Other hypotheses about creole formation have attracted attention in recent years, though not without their critics; (see Singler 1996 and Baker 2000). One is Lefebvre's 1986 relexification hypothesis, which argues that grammatical substrate influence in Haitian Creole French is so pervasive that the language should be conceptualized as “relexified” Fon (the African primary substrate source) rather than as restructured French. At the other extreme, Chaudenson 1992 attributes the features of French creoles primarily to nonstandard superstrate dialects, “basilectalized” through being learned as second languages by later slave arrivals who had less access to nonstandard French speakers than did the first generations of slaves in the Caribbean and Indian Ocean colonies. Mufwene 2001, like Chaudenson, sees strong superstratal influences in the nonstandard metropolitan dialects of French and English colonial settlers, but he allows for more subtratal influence and grammatical restructuring in the development of creoles. He also situates the evolution of these languages in a more general model of the effects of demographic and other “ecological” factors on language change (see also Parkvall 2000 on demographics).

It seems to me that the debates over Bickerton's hypothesis have somewhat exaggerated the conceptual distance between genetic and cultural evolution. Where social learning is involved, perhaps it's normal for the phenotype to emerge over multiple generations, and to involve complex relationships among genetic and social structures.

Turning back to birds, one of the earliest Language Log posts, back on 10/11/2003, was about the "Emergence of birdsong phonology" in zebra finches who hear adult models during their critical period. The cited movie, also from Ofer's lab, is available here.

And last year, I discussed some over-the-top speculation about syntax in finches and humans ("Finch phrase structure", 10/1/2007), and the unique role of (a different species of) finches in Belgian culture: "Watch out for those Wallonian finches", 5/22/2007; "Dialect variation in the terminal flourishes of Flemish chaffinches", 5/25/2007; "Finches again", 6/9/2007.

[I have to add that the Olga Fehér's zebra finch colony has been evolving for many generations in a space carefully designed to be acoustically isolated from other influences. Their home was made from a discarded refrigerator, turned on its side, appropriately instrumented and ventilated, and modified so as to turn the freezer and the vegetable crispers into darker nesting areas, with the central space retained as a sort of zebra-finch living room. I believe that if there were a prize for the best scientific use of an abandoned household appliance, Olga would be a leading candidate.]


  1. Shimon Edelman said,

    May 9, 2008 @ 8:57 am

    In connection with "creole birdsong", the concept of "the logical problem of language evolution" seems relevant. It appears in Onnis and Chater (Onnis, L. Roberts, M. & Chater, N. (2002). The logical problem of language evolution. Talk presented at the IV International conference on language evolution, Harvard University; Roberts, M., Onnis, L., & Chater, N. (2005). Language Acquisition and Language Evolution: Two puzzles for the price of one. In M. Tallerman (Ed.) Prerequisites for the evolution of language, Oxford: Oxford University Press). It is also the subject of a great new paper by Christiansen and Chater, about to appear in Behavioral and Brain Sciences (preprint available here:


  2. Phil said,

    May 9, 2008 @ 10:00 am

    The development of more complex communications doesn't necessarily imply a specialized birdsong faculty – the birds could be using more general cognitive skills. But one feature of this story might imply something like a "birdsong instinct": the fact that the development of birdsong in the isolate colony stops at the same level as in other populations.
    It could be that finches sing at the level to which their cognitive abilities allow them to progress, and all finches are pretty much the same in cognitive ability. Or it could be that finches have a specialized faculty which tends to normalize all language input to a given level. I can't readily think of a way to separate these two hypotheses.

    [myl: Let's not get carried away here. The finches' songs are not like human language in the sense of having any systematic correspondence between structure and meaning. In fact, the only meanings involved in birdsong, as far as anyone knows, are, roughly, "I love you" and "get off my land". The complex and stereotyped structure of a given species' songs don't play any role other than to identify the singer to conspecifics, and to demonstrate mastery of a physically and neurologically demanding skill.

    In discussing language learning by human children, the role of developments in general cognitive ability obviously has to be taken into account. We wouldn't consider the development of counterfactual conditionals (say) as purely a matter of when certain morphological and syntactic abilities are acquired. But in the case of songbirds, I don't think that there's any reason to think that general cognitive abilities play any important role in constraining their singing behavior. For example, one of the reasons that biologists study zebra finches is that their songs are quite simple and stereotyped, compared (for example) to Bengalese finches, much less nightingales (see the transition diagrams here for examples of what this might mean in concrete terms). But that doesn't mean that the zebra finches are more limited in general cognitive abilities. ]

  3. Rubrick said,

    May 9, 2008 @ 3:55 pm

    Another contender might be Gallileo's dropping of two toasters of unequal weight from the Tower of Pisa, but of course that's widely considered apochrypal.

  4. Derek Bickerton said,

    May 10, 2008 @ 4:30 am

    Mark, you say that "Where social learning is involved, perhaps it's normal for the phenotype to emerge over multiple generations." And you may well be right, since social learning has nothing to do with creolization. How can you "socially learn" something for which you have no model, which didn't exist until you made it?

    All the "evidence" for multiple-generation creolization is spurious. The Nicaraguan Sign Language case is sheer happenstance–it just happened that only children above 10 were recruited in the early years, We simply don't know what would have happened if much younger children had been recruited from the start, I see no logical reason why they would have needed anything beyond home sign input. Roberts's claim for a two-generation model for Hawaiian Creiole is based on bizarre and counterfactual assumptions at least some of which are contradicted by her own data.

    As for the claim that the bioprogram hypothesis is defective because it fails to account for all creoles, that's simply ridiculous. Does giving a whole lot of disparate things the same name turn them into a natural class? If you believe that, you must believe that the social matrix in which new languages are formed has absolutely no effect on the outcome. However, plantation creoles form a natural class with surprisingly similar social, economic and demographic patterns, calculated to generate maximal access to the bioprogram–which of course, rather than producing a catch-all theory of creoles, was what I was after in the first place.

    All of this is in my latest book (Bastard Tongues, Hill & Wang 2008) which is not your regular linguistics book–one reader called it the most fun book about language they'd ever read–but a no-holds-barred account of my career and adventures in four continents and innumerable islands. (See reviews in NY Times, LA Times, New Scientist, etc. etc.) It will royally piss off lots of creolists, but entertain just about anyone else. (Sorry about the ad, but in the immortal words of Truman Capote, "A boy has to hustle his book".)

  5. Mark Liberman said,

    May 10, 2008 @ 9:52 am

    Hi Derek!

    I've tried to answer your question "How can you 'socially learn' something for which you have no model, which didn't exist until you made it?" in another post ("A multi-generational bioprogram? Derek Bickerton objects").

  6. Tom Windsor said,

    May 11, 2008 @ 2:23 pm

    Perhaps we should all learn Esperanto instead. That would be even more ridiculous than pidgin English

RSS feed for comments on this post