In a comment on the post, Chris Lucas suggested that languages at higher altitudes might be more isolated, and so less subject to contact-induced change:

“contact tends to make languages lose ejectives, if they ever had them. The reasoning here would be that a language’s having (contrastive) ejectives implies that it has a large consonant inventory, which implies that it does not have a history of significant numbers of people having learnt it as a second language, since this tends to lead to the elimination of typologically rare features.”

We can test this in the following way: We can get a rough proxy for langauge contact for a community by counting the number of languages within 150km (range between 0 and 44).  If we run a phylogenetic genralised least squares test, predicting the presence of ejectives by elevation and number of surrounding languages, we get the following result (estimated lambda = 0.8169142 , df= 491, 489):

While elevation is still significant, the number of surrounding languages is also a significant predictor (the effect size is also greater).  The greater the number of surrounding languages, the smaller the chance of a langauge having a ejectives.  This fits with the idea that contact induced change removes ejectives, rather than air pressure being the only cause.   In the graph below, I’ve plotted the mean elevation for languages with and without ejectives, comparing languages with a neighbour within 150km and languages without a neighbour in 150km.  The effect is stronger in the group with neighbouring languages (right), which would fit with languages loosing ejectives due to contact.

However, it’s not quite so simple, since we have to take into account the relative relatedness of languages.  If we count the number of distinct language families within 150km, then the significance goes away:

What about another proxy for contact, like population size (as used by Lupyan & Dale, 2010)?  I took speaker populations from the Ethnologue and ran another PGLS:

Now we see that neither variable is significant, though larger populations tend not to have ejectives.  That is, by controlling for linguistic descent and population size, the correlation between elevation and ejectives goes away.

In fact, a simple logit regression predicting elevation by elevation and log population results in the following:

We can see that, even if we don’t control for phylogeny, population size is a better predictor of ejectives than elevation (although Everett uses several measures of altitude).

I also wondered if the distance to the nearest language could be a proxy for contact.  Let’s put all the variables into one regression.

Here we see that the number of surrounding languages is still a significant predictor of the presence of ejectives (although using the number of surrounding families doesn’t work), but elevation is not.

We can build the most likely causal graph (see my post here) for the data above.  This ignores the phylogenetic relatedness of langauges, but allows us to explore more complex relationships between all the variables.  Below, we see that elevation and ejectives are still linked, as Everett would predict.

The stats I’ve presented here are just rough explorations of the data, not proof or disproof of any theory.  Here are some issues that are still unresolved:

• What about the distance from high-elevation areas, as used in Everett’s paper?
• Are the proxies above reasonable?
• What is the likelihood of keeping ejectives versus losing them during contact?
• In the analyses above, I’m not controlling for geographical relatedness, this could be done by selecting independent samples or Mantel tests.
• There are links between phoneme inventory size, the geographic area a langauge covers, morphology and demography (see James’ posts here and here).  What is the best way to approach the complex relationships between these features?

Of course, laboratory experiments or careful idographic work could address these issues better than more statistics.

The first passage is from George Williams, a biologist. It’s in a chapter from a book edited by John Brockman, The Third Culture: Beyond the Scientific Revolution:

Evolutionary biologists have failed to realize that they work with two more or less incommensurable domains: that of information and that of matter. I address this problem in my 1992 book, Natural Selection: Domains, Levels, and Challenges. These two domains will never be brought together in any kind of the sense usually implied by the term “reductionism.” You can speak of galaxies and particles of dust in the same terms, because they both have mass and charge and length and width. You can’t do that with information and matter. Information doesn’t have mass or charge or length in millimeters. Likewise, matter doesn’t have bytes. You can’t measure so much gold in so many bytes. It doesn’t have redundancy, or fidelity, or any of the other descriptors we apply to information. This dearth of shared descriptors makes matter and information two separate domains of existence, which have to be discussed separately, in their own terms.

The gene is a package of information, not an object. The pattern of base pairs in a DNA molecule specifies the gene. But the DNA molecule is the medium, it’s not the message. Maintaining this distinction between the medium and the message is absolutely indispensable to clarity of thought about evolution.

Just the fact that fifteen years ago I started using a computer may have had something to do with my ideas here. The constant process of transferring information from one physical medium to another and then being able to recover that same information in the original medium brings home the separability of information and matter. In biology, when you’re talking about things like genes and genotypes and gene pools, you’re talking about information, not physical objective reality. They’re patterns.

I was also influenced by Dawkins’ “meme” concept, which refers to cultural information that influences people’s behavior. Memes, unlike genes, don’t have a single, archival kind of medium. Consider the book Don Quixote: a stack of paper with ink marks on the pages, but you could put it on a CD or a tape and turn it into sound waves for blind people. No matter what medium it’s in, it’s always the same book, the same information. This is true of everything else in the cultural realm. It can be recorded in many different media, but it’s the same meme no matter what medium it’s recorded in.

It seems to me that that is more or less how the concept of information is used in many discussions. It’s certainly how Dennett tends to use it. Here’s a typical passage (it’s the fifth and last footnote in From Typo to Thinko: When Evolution Graduated to Semantic Norms):

There is considerable debate among memeticists about whether memes should be defined as brain-structures, or as behaviors, or some other presumably well-anchored concreta, but I think the case is still overwhelming for defining memes abstractly, in terms of information worth copying (however embodied) since it is the information that determines how much design work or R and D doesn’t have to be re-done. That is why a wagon with spoked wheels carries the idea of a wagon with spoked wheels as well as any mind or brain could carry it.

Here I can’t help but think that Dennett’s pulling a fast one. Information has somehow become reified in a way that has the happy effect of relieving Dennett of the task of thinking about the actual mechanisms of cultural evolution. That in turn has the unhappy effect of draining his assertion of meaning. In what way does a wagon with spoked wheels carry any idea whatsoever, much less the idea of itself?

* * * * *

It seems to me that we can talk of a signal as carrying information only with respect to something that can read that signal (and act on what it reads), write it, or both. Information is thus a relationship between signals and mechanisms that use those signals. Both Dennett and Williams, however, seem to talk about information as though it were itty bitty particles that one can, for example, transfer “from one physical medium to another.”

But that’s not what’s going on. In the case of Don Quixote, to use Williams’s example, it takes a fairly sophisticated system to “transfer” the “information” from one medium to another. If you want to transfer the paper to text to, say, microfilm, then you have to photograph each page of the text. But it’s not at all clear to me that any information has been “transferred.” Rather, a certain pattern has been copied from medium to another well enough so that a person can read the same “information” in either version. That is, the informatic relationships between the person and the image remain the same for both versions of the image.

Things get a lot trickier when we talk of “transferring” information between a written text and a spoken text. Until relatively recently, the only way to make such transfers has been to have a person read the text, in one direction (text to speech), or transcribe it in the other direction (speech to text). One could argue that, in any event, the spoken version “contains” more “information” than the written version (vocal inflections of various sorts). In recent years we’ve developed computer programs that can perform translations on both directions, but text to speech doesn’t sound natural, and speech to text makes more errors that humans do.

* * * * *

I don’t know what to do about Dennett’s example. It’s too open-ended for analysis. It’s easy enough to take it as is, where it’s an informal statement. But I don’t see how any sophisticated argument can be built on that statement nor, for that matter, do I have any reason to believe that it’s an informal statement of an argument that’s been made in a detailed and sophisticated way somewhere else.

In the first place, we need to distinguish between using a wagon to move things from one place to another, and constructing a wagon. I can imagine, for example, that people who have never seen a spoke-wheeled wagon could use one easily enough, especially if they already have experience using wagons with solid wheels. Constructing wagons is another matter. Would people with no experience of wagons (and comparable artifacts) be able to reverse engineer a spoke-wheeled wagon given nothing more that an example of one? More plausibly, would people with considerable experience in constructing wagons with solid wheels be able to reverse engineer spoked-wheels from examples alone?

It’s not at all obvious that they would, as the construction of spoked wheels is rather sophisticated (for an analysis see F. T. Cloak, Jr., Cultural Darwinism: Natural Selection of The Spoked Wood Wheel (PDF).

Now, you may be thinking that this is all rather pedantic and picky. Picky, yes. Pedantic, no. It’s about details, and the patterns they take. Biologists know a great many details about DNA, its replication, and its role in development. Information is about those details; without them the concept is meaningless. Memeticists, I’m afraid, are short on details. What’s worse, they don’t seem to think they’re important.

More later.

Image from the JEPS Bulletin

As promised, and first thing’s first, when writing a systematic review, how should we phrase our research question? This is useful when phrasing questions for individual studies too.

PICO is a useful mnemonic for building research questions in clinical science:

• Patient group
• Intervention
• Comparison/Control group
• Outcome measures

How does this look in practice?

What is the effect of [intervention] on [outcome measure] in [patient group] (compared to [control group])?

How can we make this more applicable for language evolution?

I guess we can change the mnemonic:

Population (either whole language populations in large scale studies, small sample populations either in the real world or under a certain condition in a laboratory experiment, or a population of computational or mathematical agents or population proxy)

Intervention
Comparison/Control group
Outcome measures

Here are some examples of what this might look like using language evolution research:

What is the effect of [L2 speakers] on [morphological complexity] in [large language populations] compared to [small language populations]?

What is the effect of [speed of cultural evolution] on [the baldwin effect] in [a population of baysian agents]?

What is the effect of [iterated learning] on [the morphosyntactic structure in an artificial language] in [experimental participants]?

What is the effect of [communication] on the [distribution of vowels] in [a population of computational agents]?

All of the above are good research questions for individual studies, but I’m not sure it would be possible to do a review on any of the above research questions simply because there is not enough studies, and even when studies have investigated the same intervention and outcome measure, they haven’t used the same type of population.

In clinical research the same studies are done again and again, with the same disease, intervention and population. This makes sense as one study does not necessarily create enough evidence to risk people’s lives on the results. We don’t have this problem in language evolution (thank god), however I feel we may suffer from a lack of replication of  studies. There has been quite a lot of movement recently (see here) to make replication of psychological experiments encouraged, worthwhile and publishable. It is also relatively easy to replicate computational modelling work, but the tendency is to change the parameters or interventions to generate new (and therefore publishable) findings. And real world data is a problem because we end up analysing the same database of languages over and over again. However, I suppose controlling for things like linguistic family, and therefore treating each language family as its own study, in a way, is a sort of meta-analysis of natural replications.

I’m not sure there’s an immediate solution to the problems I’ve identified above, and I’m certainly not the first person to point them out, but thinking carefully about your research question before starting to conduct a review is very useful and excellent practice, and you should remember that when doing a systematic review, the narrower your research question, the easier, more thorough and complete your review will be.

This sounds suspiciously like one of our spurious correlations – links between cultural features that come about by accidents of cultural history rather than being causally related.  Although Everett notes that the tests he uses include languages from many language families, there’s no real control for historical descent.  James and I have also submitted a paper to PLOS ONE about this phenomenon more generally, and we suggest a few statistical tests that should be applied to this kind of claim.  These include comparing the correlation of the variables of interest with similar variables that you don’t think are related, and controlling for historical descent by using, for example, phylogenetic generalised least squares.  In this post, I apply these tests.

First, I test whether the link between ejectives and elevation is stronger than the link between elevation and many other linguistic features.  I ran a correlation for each variable in the WALS database.  Elevation (altitude) does indeed significantly predict the presence of ejectives.  Surprisingly, only 2 other variables resulted in stronger predictors of elevation.  That is, the presence of ejectives is in the top 1.4% of variables for predicting elevation.  The presence of ejectives resulted in a correlation that was significantly stronger than 94.4% of variables (above 1.98 standard deviations). This is surprisingly good news for Everett!

Below is a histogram of the results (F-score of the model fit), with a red line indicating the strength of the ejectives variable :

The linguistic variables that gave better results than ejectives were the Order of Object and Verb and the Relationship between the Order of Object and Verb and the Order of Adjective and Noun. I can’t think of a good reason that these would be linked.  See below:

The next test involved controlling for common descent of languages.  I built a phylogenetic tree from the linguistic classifications from the Ethnologue.  We’re predicting elevation (continuous) given the presence of ejectives (discrete), so we’ll use a phylogenetic generalised least squares test (you can learn more about doing this at the excellent tutorials by Charles Nunn and others, here).  This weights the observations by how related they are, given a particular model of trait evolution.  The elevation variable has a strong phylogenetic signal (Pagel’s lambda = 0.3, sig. > 0, p<0.00001; sig. different from 1, p<0.00001), so we’ll use Pagel’s covarience matrix.

Surprisingly, the correlation holds up, even when controlling for phylogeny (491 languages, df = 419, residual df = 489, estimated lambda = 0.2787271, coef = 358.9542, t = 3.51, p = 0.0005).  Edit: If you use ejectives as the dependent variable, the result is similar (estimated lambda = 0.8169142, coef = 0.00003975, t = 2.42, p = 0.0157).

I’d like to make two points:  First, this kind of analysis is easy to do, and makes the test more rigorous (I did the above analyses at Singapore airport).  Secondly, while the stats might hold up, this kind of approach can only point towards future research, rather than supplying definitive proof of the hypothesis.  It’s an interesting proposal, and I look forwards to some modelling or experimental evidence.

EDIT:

The phylogenetic tree assumed languages within families evolved over 6,000 years and there was a common ancestor for all language families 60,000 years ago. You can see a diagram of the tree here, with WALS codes.

The altitude data I used comes from the 90-meter NASA database (SRTM3), extracted using the GPS Visualiser, while Everett uses surveys by Google Earth and ArcGIS v. 10.0.  I checked some points and there are very slight differences in the order of a few meters.

So first thing’s first, what is a systematic review? A systematic review is defined (by the Centre for Reviews and Dissemination at the University of York) as “a review of the evidence on a clearly formulated question that uses systematic and explicit methods to identify, select and critically appraise relevant primary research, and to extract and analyse data from the studies that are included in the review.”

This is in contrast to more narrative or literature reviews, more traditionally seen in non-medical disciplines. Reviews within language evolution are usually authored by the main players in the field and are generally on a very broad topic, they use informal, unsystematic and subjective methods to search for, collect and interpret information, which is often summarised with a specific hypothesis in mind, and without critical appraisal, and summarised with an accompanying convenient narrative. Though these narrative reviews are often conducted by people with expert knowledge of their field, it may be the case that this expertise and experience may bias the authors. Narrative reviews are, by definition, arguably not objective in assessing the literature and evidence, and therefore not good science. Some are obviously more guilty than others, and I’ll let you come up with some good examples in the comments.

So how does one go about starting a systematic review, either as a stand alone paper or as part of a wider thesis?

Systematic reviews require the following steps:

From: YourHealthNet in Australia

1. Phrase the research question

2. Define in- and exclusion criteria (for literature search)

3. Search systematically for all original papers

4. Select relevant papers

5. Assess study quality and validity

6. Extract data

7. Analyse data (with a meta-analysis if possible)

8. Interpret and present data

In the coming weeks I will write posts on how to phrase the research question of your review, tips on searching systematically for relevant studies, how to assess the quality and validity of the papers and studies you wish to cover, and then maybe a post on meta-analysis (though this is going to be difficult with reference to language evolution because of its multidisciplinary nature and diversity within the relevant evidence, I’ll have a good think about it)

References

Undertaking Systematic Reviews of Research on Effectiveness. CRD’s Guidance for those Carrying Out or Commissioning Reviews. CRD Report Number 4 (2nd Edition). NHS Centre forReviews and Dissemination, University of York. March 2001.

This one, by Rafferty, Griffiths & Ettlinger, to appear in Cognition, uses an iterated learning experiment to challenge the idea that tendencies across cultures  is the result of some structures and concepts being easier to learn than others, as things being easier to learn means they will be more accurately transmitted from one generation to the next. Mini artificial languages in iterated paradigms (most notably Kirby, Cornish & Smith, 2008), have shown that languages become more structured as the result of generational turnover (and with an added pressure for expressivity), and this is hypothesised to be because of pressures for learnability (as well as expressivity/communication).

If we can show empirically that cultural features which are more prevalent are more “learnable”, than this adds extra weight to the hypothesis that the driving force because culturally universal concepts are the result of learnability. However, this paper finds the opposite, if a concept is more learnable, then that does not necessarily result in it being more prevalent in transmission chains.

Their first argument is that more learnable cultural features are not likely to be (re)produced in transmission failure. This was shown in an experiment which featured “distinctive items”, such as the word “Elephant” on a shopping list. In this context, the word “Elephant” was much more likely to be remembered than other items on a list, but once it had been lost in a transmission chain, it was never regenerated. Participants were much more likely to regenerate mundane food items which are likely to feature on a shopping list, such as “apple”.

They also showed this mathematically, showing that agents are more likely to arrive at H2 if they learn from an agent with H2, even if H1 is more learnable. This is based on the assumptions that learners rarely learn a particular hypothesis unless they receive data generated specifically from that hypothesis, less learnable hypotheses are more likely to be confused with one another and so will arise more often through transmission, and that learnable hypotheses are unlikely to arise as the result of transmission errors, just like the word “elephant”.

The main point of this isn’t to argue that learnability doesn’t play a role in cultural transmission, but rather that we shouldn’t look at only learnability bias, because we also need to take into account the rate at which hypotheses (or cultural features) change into other hypotheses.

This is quite a neat experiment, and works well in explaining certain cultural trends we see. The paper cites the example that across religions, concepts are generally “minimally counterintuitive” (Boyer, 1994). This is often cited to be because things that are minimally counterintuitive are more learnable, because counterintuitive elements are learnable because they are surprising, but if you have too many surprising things in a story, people can’t remember all of them and so they can’t be transmitted. Minimally counterintuitive stories have an optimal amount of counterintuitive elements to be memorable and so are transmitted with more fidelity (see Norenzayan et al. 2006 for a really neat study). However, Rafferty. Griffiths & Ettlinger show that more than just learnability needs to be taken into account because even one counterintuitive element is unlikely to be reinvented if forgotten. A good linguistic example of this is clicks, clicks have been shown to be the most acoustically salient sounds, but they appear in very few languages. This could be because they are unlikely to be spontaneously (re)produced, so are lost despite their learnability.

The scope of the implications for this study are limited however, as it makes the massive assumption that learnability is only born from an item being surprising. Sure, if all you’re learning is a shopping list, the most learnable thing is the elephant, because it is the most surprising, but if you’re learning a whole language, the surprising things are not going to facilitate learning an entire lexicon and grammar. In the context of language learning, structure is the thing that makes it more easily learned, rather than things being surprising.

The paper then goes on to test whether, if a set of languages that lack a property far outnumber the set with a property, then this can override a learnability bias. They carry out an individual artificial language learning experiment which shows that the learnability and generalisability of words with vowel harmony is greater than words without vowel harmony. However, they then go on to do an iterated learning experiment and find that no matter how much of an initial language consists of words with vowel harmony, all languages ended up with approximately 50% words with vowel harmony. This shows that, while one language is more accurately transmitted than others in an individual learning task (or in one generation), because of the large number of non-harmonic possibilities, a completely harmonic language can not win out through transmission.

There’s quite an extensive paragraph explaining why vowel harmonic languages should exist at all then:

In contrast with the results of our experiment, harmony does exist in many languages of the world. Several factors might result in harmony being more common in these languages than in the final generations of our chains. Our experiments focus on cognitive learning biases, but it is likely that there are also sensorimotor biases that favor the articulation and perception of harmonic languages (Blevins, 2004). There may also be qualitative factors not included in our experiment that lead to the harmony bias being stronger in natural language than in the lab. For instance, children could have stronger harmony biases. Additionally, the quantitative bias towards harmony may be stronger than we found in Experiment 1. This could occur due to the existence of more words and a longer period of learning and use in naturalistic settings. Since all of our participantsLearnability and cultural universals 26 were adult speakers of English, their bias could also be weaker due to the fact that English is not a vowel-harmonic language. Another factor that could lead to a divergence between our results and natural language learning is the use of a linear transmission structure. In natural language learning, children may learn from people who are part of generations other than the prior generation. Transmission patterns are likely also influenced by other factors, such as language contact. This can result in speakers borrowing phenomena from other languages, resulting in the spread of properties that are unlikely to be generated spontaneously. Finally, there may be increased noise in transmission in lab experiments due to the fact that learning occurs over a relatively short period. This might mean that we would expect harmonic languages to eventually become less prevalent due to transmission errors, but that this process will be much slower in natural language than in the lab. In the experiment, we see that by the third generation the languages no longer contain more harmonic words than would be expected by random chance. This rapid shift may indicate that participants exhibit very little bias towards harmonic words when the input language is not 100% harmonic; in a naturalistic context, generalization is likely to be somewhat more robust due to the longer learning period and broader exposure to the language. Despite these differences, our experiment provides evidence for the fact that a bias need not lead to a universal tendency, something which is born out in the pattern of vowel harmony in existing languages: vowel harmony is relatively common, but it is not present in the majority of languages.

The conclusion is that there are a whole host of reasons why a language or cultural phenomenon might be successful in being transmitted and the learnability isn’t necessarily enough. But I’m not sure anyone was arguing it was in the first place.

References

Boyer, P. (1994). The naturalness of religious ideas: A cognitive theory of religion. Univ of California Press.

Kirby, S., Cornish, H., & Smith, K. (2008). Cumulative cultural evolution in the laboratory: An experimental approach to the origins of structure in human language. Proceedings of the National Academy of Sciences, 105(31), 10681-10686.

Norenzayan A, Atran S, Faulkner J, & Schaller M (2006). Memory and mystery: the cultural selection of minimally counterintuitive narratives. Cognitive science, 30 (3), 531-53 PMID: 21702824

Rafferty, A. N., Griffiths, T. L., & Ettlinger, M. (in press). Greater learnability is not sufficient to produce cultural universals. Cognition.

In the first piece Dennett seems to be using the standard-issue computational model/metaphor that he’s been using for decades, as have others. This is the notion of a so-called von Neumann machine with a single processor and a multi-layer top-down software architecture. In the second and more recent piece Dennett begins by asserting that, no, that’s not how the brain works, I was wrong. At the very end I suggest that the idea of the homuncular meme may have served Dennett as a bridge from the older to the more recent conception.

Words, Applets, and the Digital Computer

As everyone knows, Richard Dawkins coined the term “meme” as the cultural analogue to the biological gene, or alternatively, a virus. Dennett has been one of the most enthusiastic academic proponents of this idea. In his 2009 Cold Spring Harbor piece Dennett concentrates his attention on words as memes, perhaps the most important class of memes. Midway through the paper tells us that “Words are not just like software viruses; they are software viruses, a fact that emerges quite uncontroversially once we adjust our understanding of computation and software.”

Those first two phrases, before the comma, assert a strong identification between words and software viruses. They are the same (kind of) thing. Then Dennett backs off. They are the same, providing of course, that “we adjust our understanding of computation and software.” Just how much adjusting is Dennett going to ask us to do?

This is made easier for our imaginations by the recent development of Java, the software language that can “run on any platform” and hence has moved to something like fixation in the ecology of the Internet. The intelligent composer of Java applets (small programs that are downloaded and run on individual computers attached to the Internet) does not need to know the hardware or operating system (Mac, PC, Linux, . . .) of the host computer because each computer downloads a Java Virtual Machine (JVM), designed to translate automatically between Java and the hardware, whatever it is.

The “platform” on which words “run” is, of course, the human brain, about which Dennett says nothing beyond asserting that it is there (a bit later). If you have some problems about the resemblance between brains and digital computers, Dennett is not going to say anything that will help you. What he does say, however, is interesting.

Notice that he refers to “the intelligent composer of Java applets.” That is, the programmer who writes those applets. Dennett knows, and will assert later on, that words are not “composed” in that way. They just happen in the normal course of language use in a community. In that respect, words are quite different from Java applets. Words ARE NOT explicitly designed; Java applets ARE. Those Java applets seem to have replaced computer viruses in Dennett’s exposition, for he never again refers to them, though they figured emphatically in the topic sentence of this paragraph.

The JVM is “transparent” (users seldom if ever encounter it or even suspect its existence), automatically revised as needed, and (relatively) safe; it will not permit rogue software variants to commandeer your computer.

Computer viruses, depending on their purpose, may also be “transparent” to users, but, unlike Java applets, they may also commandeer your computer. And that’s not nice. Earlier Dennett had said:

Our paradigmatic memes, words, would seem to be mutualists par excellence, because language is so obviously useful, but we can bear in mind the possibility that some words may, for one reason or another, flourish despite their deleterious effects on this utility.

Perhaps that’s one reason Dennett abandoned his talk of computer viruses in favor of those generally helpful Java applets.

But then why would he have talked about computer viruses in the first place? Simple: tradition. Memetics talk has long used the notion of a (cultural) virus either as an alternative to or in alternation with the notion of a (cultural) gene; and memetics has talked of computer viruses for some time now. It’s a useful analogy.

Now that he has banished computer viruses, and their nasty effects, from our minds, Dennett then goes on to assert:

Similarly, when you acquire a language, you install, without realizing it, a Virtual Machine that enables others to send you not just data, but other virtual machines, without their needing to know anything about how your brain works.

Language acquisition and learning has now become a matter of installation. That, it seems to me, marks quite a difference between computer technology and natural language. Software installation is a relatively quick and straightforward process and is something that a human agent does to a computer. Language learning takes place over a decade or so and is primarily self-directed but with external assistance by others. Similarly, it is easy to uninstall a piece of software. But how would you “uninstall” someone’s knowledge of a language? One might well do so by destroying a large part of their brain, but that would likely destroy much else as well. Knowledge and skills reside in brains in a way that is much different from how software exists in computers.

Dennett of course knows that language learning is not the same as software installation and talks about it later in the paper. But he firsts re-establishes the parallel between words and software:

Words are not just sounds or shapes. As Jackendoff (2002) demonstrates, they are autonomous, semi-independent informational structures, with multiple roles in cognition. They are, in other words, software structures, like Java applets.

Dennett then goes on to distinguish between intelligent design (of Java applets) and blind evolution (of language) and he tells us something about how words are installed:

Unlike Java applets, they are designed by blind evolution, not intelligent designers, and they get installed by repetition, either by deliberate rehearsal or via several chance encounters. The first time a child hears a new word, it may scarcely register at all, attracting no attention and provoking no rehearsal; the second time the child hears the word, it may be consciously recognized as somewhat familiar or it may not, and in either case, its perception will begin laying down information about context, about pronunciation, and even about meaning.

The paragraph goes on to say a bit more about language acquisition and to note, following Terrence Deacon, that the structure of the brain must have co-evolved with the emergence of language.

I want to dwell on these two differences:

1) Intelligent and deliberate design of software vs. “design” of language through blind evolution.

2) Human installation of software in a computer as a relatively quick ‘one-shot’ process vs. language acquisition and learning by a child over a period of years.

These are two aspects of the same thing, the fact that computer systems, software and hardware, are designed by humans who have a comprehensive and external overview of those systems while language evolves in humans who do not have a comprehensive and external overview of it.

External Designers vs. Neurons as Agents

As my friend and colleague Tim Perper put it years ago, the trouble with the analogy between computers and brains is that computers are designed and programmed by humans but brains are not. While I agreed with Tim–that observation, after all, has been around for some time–I also thought to myself that that difference was somehow external to the comparison. And for some purposes perhaps it is; I know I’ve certainly found it useful to think of the mind as being computational in some important way. But more and more I find myself agreeing with Tim, that we can’t simply ignore that difference. It is fundamental. It is not so much about the general idea of computing, but about a specific style of computing.

Dennett, of course, does not deny a difference between brains and computers. It’s there in the care he takes to differentiate the intelligent design of Java applets from the evolutionary design of words. But he does no more than state the difference. He makes no attempt to probe it nor does he suggest that it might undercut the force of his analogy, between words and applets.

Now, let’s take a look at his recent interview, The Normal Well-Tempered Mind. The interview is informal unlike his Cold Spring Harbor article. Dennett opens with a moderately dramatic statement:

I’m trying to undo a mistake I made some years ago, and rethink the idea that the way to understand the mind is to take it apart into simpler minds and then take those apart into still simpler minds until you get down to minds that can be replaced by a machine.

So, he made a mistake years ago–near the beginning of his career I would think–and he’s trying to think himself out from under it. Good enough. He goes on to say:

The idea is basically right, but when I first conceived of it, I made a big mistake. I was at that point enamored of the McCulloch-Pitts logical neuron. McCulloch and Pitts had put together the idea of a very simple artificial neuron, a computational neuron, which had multiple inputs and a single branching output and a threshold for firing, and the inputs were either inhibitory or excitatory. They proved that in principle a neural net made of these logical neurons could compute anything you wanted to compute.

McCulloch and Pitts and did this work in the 1940s and it was common lore, at least in some circles, by the time Dennett went to college and graduate school (you can find the important papers in Warren S. McCulloch, Embodiements of Mind 1965). Dennett goes on to say:

So this was very exciting. It meant that basically you could treat the brain as a computer and treat the neuron as a sort of basic switching element in the computer, and that was certainly an inspiring over-simplification. Everybody knew is was an over-simplification, but people didn’t realize how much, and more recently it’s become clear to me that it’s a dramatic over-simplification, because each neuron, far from being a simple logical switch, is a little agent with an agenda, and they are much more autonomous and much more interesting than any switch.

The computers on which one can install a Java Virtual Machine, and for which one can write a variety of Java applets, ARE constituted of millions of logical switches, realized in silicon. These circuits, of course, are designed by engineers, as are the layers of software that runs in those circuits.

But the neurons that constitute human brains ARE NOT simple switches. They are, as Dennett now emphasizes, more or less autonomous agents. When language is being acquired, it is being acquired by billions of such agents, connected together in a brain, pursuing their various individual agendas even as the “owner” of that brain is living her life.

Dennett goes on to ask:

The question is, what happens to your ideas about computational architecture when you think of individual neurons not as dutiful slaves or as simple machines but as agents that have to be kept in line and that have to be properly rewarded and that can form coalitions and cabals and organizations and alliances? This vision of the brain as a sort of social arena of politically warring forces seems like sort of an amusing fantasy at first, but is now becoming something that I take more and more seriously, and it’s fed by a lot of different currents.

And then:

First, you try to make minds as simple as possible. You make them as much like digital computers, as much like von Neumann machines, as possible. It doesn’t work. Now, we know why it doesn’t work pretty well. So you’re going to have a parallel architecture because, after all, the brain is obviously massively parallel.

It’s going to be a connectionist network. Although we know many of the talents of connectionist networks, how do you knit them together into one big fabric that can do all the things minds do? Who’s in charge? What kind of control system? Control is the real key, and you begin to realize that control in brains is very different from control in computers. Control in your commercial computer is very much a carefully designed top-down thing.

The topmost level of control over digital computers is external to the machines themselves. It is the prerogative on the humans who design, build, program, and operate the machines. In contrast, human brains and human language have evolved without such external control. They have somehow arisen in populations of interacting neurons, each of them active agents.

Dennett goes on from there to talk about culture as providing “opportunities that don’t exist for any other brain tissues in any other creatures, and that this exploration of this space of cultural possibility is what we need to do to explain how the mind works.” And so:

My next major project will be trying to take another hard look at cultural evolution and look at the different views of it and see if I can achieve a sort of bird’s eye view and establish what role, if any, is there for memes or something like memes and what are the other forces that are operating. We are going to have to have a proper scientific perspective on cultural change.

It’s a worthy project. Now that he’s abandoned the idea of neurons as simple switches and has started to think about them as active agents, will he abandon the metaphorical and analogical use of computing technology based on such simple switches?

From Switches Through Memes to Neural Agents?

What I find particularly interesting and curious in Dennett’s thinking as it is exhibited in these two pieces is the relative absence of talk about memes as agents. For he has certainly talked in that way and even endowed those memetic agents with the power to inculcate irrational beliefs in otherwise rational human beings. What I’m wondering is if, in effect, such agency has become detached in Dennett’s mind from the concept of memes flitting about from one brain to another and has now become lodged in those complex neurons “that can form coalitions and cabals and organizations and alliances.” I note, in parting, that the same McCulloch who gave us the neuron as logical switch also gave us a concept of competitive control in which the ultimate control of the brain is lodged, not at the top, but at the bottom, in the reticular activating system.

* * * * *

If you want to play around with the idea of the brain as consisting of millions of quasi-autonomous agents, see my post, The Busy Bee Brain. For a brief and informal introduction of McCulloch’s model of the reticular activating system, see Mode & Behavior 2: McCulloch’s Model.

Both Hannah and I are at the conference and we’re also live-tweeting about the conference using the hashtag #protolang3

Hannah’s just given her talk

Jack J. Wilson, Hannah Little (University of Leeds, UK; Vrije Universiteit Brussel, Belgium) – Emerging languages in esoteric and exoteric niches: evidence from rural sign languages (abstract here)

And I’m due tomorrow.

Michael Pleyer (Heidelberg University, Germany) - Cooperation and constructions: looking at the evolution of language from a usage-based and construction grammar perspective (abstract here)

The Programme can be found here: (Day 1 / Day 2)

But we’re getting ahead of ourselves. We need a little conceptual equipment before considering the example. It’s the conceptual equipment that’s in question. Make no mistake, the concept of memes is conceptual equipment, and it’s confused and confusing.

Roles in Cultural Selection

Genes and phenotypes play certain roles in a more or less standard account of biological evolution. The phenotype interacts with the environment, where it either succeeds or fails at reproduction, depending on the “fit” between its traits and that environment. Where the phenotype is successful at reproduction, it is the genes which are said to carry heredity from one generation to the next.

In one very widespread account genes are said to be replicators. That is to say, replication is the role they play in evolutionary change. Here’s what Peter Godfrey-Smith has to say about that (The Replicator in Retrospect, Biology and Philosophy 15 (2000): 403-423.):

In The Selfish Gene (1976), Richard Dawkins had argued that individual genes must be seen as the units of selection in evolutionary processes within sexual populations. This is primarily because the other possible candidates, notably whole organisms and groups, do not “replicate.” Organisms and groups are ephemeral, like clouds in the sky or dust storms in the desert. Only a replicator, which can figure in selective processes over many generations, can be a unit of selection.

At the same time Dawkins coined the term “meme” to name entities filling the replicator role in cultural evolution. Later on he used the term “vehicle” to designate the entity that interacts with the environment. In biological evolution it is phenotypes that are the vehicles. In cultural evolution, well, that’s a matter of some dispute. And that more general dispute–what are the roles in cultural evolution and what kinds of things occupy them?–is what interests me.

However, I don’t particularly like the term “vehicle.” As Godfrey-Smith has noted, following others, it is a gene-centric term, characterizing what entities do from the so-called “gene’s eye” perspective. I’d prefer a more neutral perspective and so will use a term coined by Richard Hull, “interactor.” Here are definitions as Godfrey-Smith gives them:

Replicator: an entity that passes on its structure largely intact in successive replications.
Interactor: an entity that interacts as a cohesive whole with its environment in such a way that this interaction causes replication to be differential.

We need one more role, that of beneficiary, as defined by Elisabeth Lloyd. “The beneficiary, for Lloyd, is the entity that ‘ultimately benefits’ from a process of evolution by selection” (Godfrey-Smith, see also Lloyd’s treatment of Units and Levels of Selection HERE). Godfrey-Smith goes on to suggest that “that most of the language of … ‘ultimate benefit’ in this context is merely metaphorical.” That may be so, however, I find it useful, at least heuristically.

Consider for example gene-culture coevolution school of thinking, which offers a technically sophisticated treatment of cultural change. In that tradition it is biological organisms, mostly humans, but also chimpanzees, songbirds and some other animal species, that occupy the beneficiary role. But that is not the case in any version of memetics, where it is the memes that are the beneficiaries. In fact, part of the appeal of memetics has been that it offers a way of thinking about cultural practices, such as life-long celibacy, that are at odds with biological interests. While I’ve got problems with memetics, I am very much interested in accounts of cultural evolution it which it is the cultural entity that benefits from cultural evolution.

So, we’ve got three roles on the table: interactor, replicator, and beneficiary. Let’s return to some jazz club in Philadelphia on some night in 1952 and talk about that performance of “Dexterity.”

Late Night Dexterity

The environment in which any cultural practice thrives or dies is, of course, the social group. The group relevant to the imaginary jazz performance I invoked at the beginning is, first of all, the people present in the club to witness it, the audience and, of course the musicians themselves, not to mention the other people working at the club. The jazz audience is, of course, larger and more diffuse than the people in a given club on a given occasion. But we need not worry about the demographic details, not for the purposes of this example. All I want to do is to explicitly state that the evolutionary environment for music is some population of people.

And it is the performance that plays the interactor role, not the tune, in this case “Dexterity,” but the performance of the tune. It is the performance itself that the musicians and audience will find to be satisfactory or not. The tune is simply a source musical materials on which a performance can be based.

What, then, is the beneficiary of the performance? Well, one would think it is those musical materials in the tune, “Dexterity” in this case. That’s not all, but let’s start with that.

Of what does the tune “Dexterity” consist? The conventions of jazz have it that such tunes consist of a melody and a harmonic structure. A melody is a sequence of musical tones having specific pitches and durations. Melodies are concrete and, in principle, one can hum, whistle, or sing them. In practice, that might be difficult. “Dexterity” is one such difficult case as it is fast and complex, a characteristic of many bebop tunes—bebop, of course, is a particular jazz style and “Dexterity” is a tune in that style.

A harmonic structure is more abstract than a melody and can be realized equally well in many different ways, ways that will sound different in matters of detail. Instruments that can play only one note at a time, such as trumpet or saxophone, can play melodies that are consistent with a given harmonic structure but they have trouble conveying the structure directly. Multi-note instruments, such as piano or guitar, can convey harmonic structure more directly. Within the jazz world harmonic structures are generally notated as short strings of alphanumeric characters, each of which indicates a single chord: B7, Am, EMaj9, etc.

The harmonic structure of “Dexterity”¬–and this is why I chose it as an example–happens to derive from another well-known song, George Gershwin’s “I Got Rhythm.” There are so many songs that derive their harmonic structure from “I Got Rhythm”, especially in bebop, that “Rhythm Changes” has become a term of art within the jazz world. Obviously enough it designates the harmonic structure (chord changes) of “I Got Rhythm” considered as an autonomous musical entity (for a little history, see Cultural Evolution 4: Rhythm Changes 2).

So, a successful performance of “Dexterity” will benefit the specific melody of the tune, increasing that likelihood that audience members will want to hear it again and that the musicians will perform it again. It will also benefit the specific harmonic structure, which happens to derive from another tune, “I Got Rhythm”. The “Dexterity” melody is specific in a way that satisfies the definition of replicator, and I suppose that’s true of the harmonic structure–Rhythm Changes–as well. That’s how I treated it a couple of years ago, in Cultural Evolution 4: Rhythm Changes 1, where I talked of it as being a memetic entity. Still, it’s abstract nature continues to give me pause.

But there’s more to a performance than the materials provided by the melody and chord changes. I’m imagining that this specific performance is by a quintet consisting of trumpet, tenor sax, piano, bass, and drums¬¬–a standard line-up. In this performance the group plays the melody once at the beginning and once at the end. In between we have improvised solos. “Dexterity” provides the opening and closing melody and it provides the harmonic framework for all the improvisations. More materials are necessary.

The drummer needs to play appropriate patterns throughout. The same is true for the piano player and the bass player, who have to supply accompanying materials throughout the performance. Things are a bit different for the sax player and the trumpet player, who are responsible for playing the melody at the beginning and the end. When they solo, however, they’re on their own.

A good many of these bits and pieces–variously called riffs or licks–may properly be considered to be replicators. Some of them will be standard throughout the bebop community, if not jazz in general, and so are played by many musicians and on many different tunes. But some of them may be specific to these musicians, and even to this specific performance of “Dexterity.” Perhaps the piano player, for example, comes up with a new lick that she likes a lot and that the audience likes as well. So she uses that lick in subsequent performances, not only of “Dexterity”, but of other tunes. One of those performances gets recorded and other musicians here it and take up that lick. Within two or three years it’s spread throughout the bebop community (and 40 years later gets sampled by a hip hop artist). It’s become a meme.

It is unlikely, however, that we’ll be able to analyze the whole performance as but a complex collection of replicators (that is, memes). For one thing, there’s likely to be at least some specific phrases that aren’t derived from prior materials. For another thing, the way all these derived and novel elements fit together is important. A performance, the interactor, is not a collection of replicators and other stuff joined in random order. The order is important.

Thus I do not think we can reduce the interactor (the musical performance) to a collection of replicators and other stuff, any more than we can reduce a biological organism to a collection of genes. In biology there is a developmental process in which phenotypes emerge through a process in which genes interact with an environment. In music there is a performance process, if you will, that accomplishes an analogous task.

Not only do the musicians have to perform the music, but members of the audience must do so as well. The process of listening to music is not passive in the way that recording devices passively record sounds. Members of the audience take the sounds that they hear and construct connections between them. Bits a piece of this process may be explicitly conscious, such as the aha! when you notice that that sax player has just quoted the theme song from the Woody Woodpecker cartoons (which also is based on Rhythm Changes), but most of it is unconscious. It just happens.

Regardless of one’s level of performance skills, one has to learn to hear music. The audience for our imaginary performance is likely to include some people with sophisticated performance skills, people quite capable of performing “Dexterity.” But most probably have few performance skills; they may sing in the shower or in church but otherwise they don’t perform any music. Still, they had to learn how to hear jazz.

Some who is not particularly knowledgeable abou bebop might well take great pleasure in a good performance of “Dexterity” even if they don’t quite know what’s going on. They might not know, for example that the harmonic structure is that of “I Got Rhythm.” And they might not know that that structure is 32 bars long and divided into four sections of eight bars each. But there’s something going on that they like, to they applaud loudly and seek out other performances, by the same musicians, of the same tune, and in the same idiom.

Memes: Internal or External?

Where are these replicators, these memes, to use Dawkins’s term? In a standard view within memetics (e.g. Daniel Dennett’s view) the memes are patterns in people’s brains. But that isn’t the view I took in the above discussion. The melody for “Dexterity” is a concrete physical entity out there in the physical world where, not only can it be heard by people, but it can be picked up by recording equipment. The same is true of all those licks and riffs and even of the rather more abstract harmonic structure (which I’ve argued in some detail).

If these things didn’t exist in the external world as physical objects or properties of physical objects, they would be of no use to musicians. In order to perform together musicians need to coordinate their actions, and do so in a very exacting way, one whose temporal characteristics must be measured in milliseconds. The only means they have of doing this is to listen to the sound. If the memes cannot be identified with properties of the sound stream they cannot play a role in coordinating the musician’s interaction, nor can they be conveyed to an audience.

The memes are those properties of the sound stream through which the musicians couple their actions into a single coordinated activity. The coupling extends to the audience and you can see and hear in the way they tap their feet, gasp a particularly audacious lick, stop chattering at the performance builds, and so forth.

That is, musical memes perform a coupling function during performance and it is these coupling functions that must be learned from one person to another if the individual memes are to earn a long-term place in the musical repertoire.

I’m not denying that, when a musician learns the melody of “Dexterity” that something happens in the musician’s brain. But it’s not at all obvious to me that we accomplish anything useful by saying that that something has been replicated from the brain of another musician or musicians. And the same is true in the case of audience members. In order to appreciate the performance they must already know the performance conventions to some degree or another.

I see nothing to be gained by saying musical replication is a matter of transferring replicators between brains. Very obviously there IS replication of sounds. Without that there is nothing. It is the sounds that are the replicators, not the neural patterns in brains.

What then, is the biological analog of whatever it is that’s taking place in human brains during musical performances? Ontogenetic development. In the case of single-celled organisms that’s a relatively simple process of cell division, but it’s not a null process. Not only must the genetic material be replicated, but cell structures and organs must be created. In the case of multi-celled organisms the developmental process can be quite complex.

The same is true in music. Using the melody of “Twinkle, Twinkle, Little Star” as the basis for a performance can be a relatively simple and straightforward process. But a skilled improviser could create a very sophisticated piece of music on that material. Wolfgang Amadeus Mozart was such an improviser, though he worked in a different idiom. We don’t really know whether or not he gave improvised performances of that tune, but the fact that, he composed a sophisticated set of variations on is suggests that he probably did.

That, of course, raises the question of more or less richly annotated compositions, such as those in Western classical music. But I’ve done enough for one blog post. We can consider that question some other time.

* * * * *

Addendum: When looking for memes in cultural evolution one can proceed under one of two assumptions: 1) Memes are a new kind of entity not yet identified in the human sciences. 2) Memes are entities that are already known but that have not heretofore been assigned to that role in the process of cultural evolution. I have taken the second position.

Addendum 2 (10 July 2013): In the above I’ve assigned the musical performance to the interactor role. But what do I mean by the performance? The complex sonic object that can, for example, be recorded? Or do I mean the (collective) neural event supported by that complex sonic object? The latter is what I had in mind in Chapter 8 of Beethoven’s Anvil where I asserted that “the phenotypic role in music’s evolution is played by performance-level attractors” (p. 192). Though a bit difficult to grasp, that’s what I mean; it’s the performance-level attractor that plays the interactor role.

In contrast, Levinson suggested that we should be moving away from the picture of the modal individual with a fixed language architecture.  Instead, we should embrace population thinking and recognise the variation inherent at every level of language from typology to processing and brain structures.  While languages are constrained by the processing structures of the brain, these processing structures are plastic and adapt to the language and cultures in which they are embedded.  Adults lose the ability to distinguish sounds that are not part of their language.  Recent work on linguistic planning using eye-tracking shows that the elements of a scene that speakers attend to before starting to speak differs with the canonical word order of their language.  More fundamentally, brain structures can be affected by cultural experience, such as bilingualism or singing (indeed, the effect of bilingualism on processing shows that variation itself is a fundamental constraint).  So, brains do constrain learning and processing, but are themselves subject to constraints from interaction between individuals.  Brains also change over evolutionary time, adapting to a range of pressures.  Therefore, there is a complex ecology of systems that co-evolve to define the constraints on language, and understanding these systems requires focussing on diversity.

Hagoort conceded that there was impressive variation at each level, but wondered what was meant by “fundamental” differences.  For instance, how important is the precise neural architecture of an individual?  Even within the variation pointed out, complex linguistic processing isn’t being done in the thalamus, and this is a constraint that sets a boundary on variation.  Hagoort might have pointed out that, if there was so much variation between individuals, how do they communicate so effectively and how does basic interaction happen so easily between diverse individuals?  This points to brain processing universals that explain the constraints on language.

Both sides agreed that the basic aim of any science, including linguistics, is to discover general principles that explain the data.  However, are researchers focussing on the same data?  What is the object of study that linguistics are trying to find generalisations for?  It seems to me that the debate came down to what each proponent thought was the domain that was most likely to yield general explanations.  Hagoort suggests that we should be focussed on brain structures and processing in the individual.  Levinson, on the other hand, suggests that the interaction between individuals is a key domain (e.g. the interaction engine).  Proponents of cultural evolution such as Simon Kirby might argue that cultural transmission is a key domain.  It’s possible that the most relevant ‘universals’ in each of these domains may be very different.  A constructive step would be to describe how each of these domains constrain the other.  For instance, constraints on language processing in the brain certainly constrain interaction between individuals, but the requirements of interaction may affect how processing is employed.

There were some good points from the floor, including Peter Seuren pointing out that neither view was particularly close to proving their point, since proving universals, or their absence is very difficult.  A paper under review by Steven Piantadosi and Edward Gibson attempts to answer whether it is possible in principle or practice to amass sufficient evidence for a statistical test that would demonstrate a universal.  They conclude that it is possible in principle, but that there are not enough datapoints (languages) in order to achieve the required statistical power.  There was also an appeal for the study of diversity for the sake of diversity – that there are different motivations for explaining phenomena in the world, and that one of them is to understand human diversity.

The general message:  Proponents of universals need to take diversity into account, and proponents of diversity need to be more specific about how diversity maps onto processing and how different domains of language co-evolve.