The Language Evolution Tree

According to one theory of language evolution, structure emerged gradually, with sub-parts of words being re-interpreted by learners as referring to sub-parts of the things they referred to. However, for this process to work, there must be some chance similarities between words and referents to begin with. This has been a major criticism of the theory, but here I present some evidence that such chance coincidences do occur (tipped off by a colleague).

Below are the front covers of two books on language evolution. The first is Christiansen & Kirby's Language Evolution, the second is Tecumseh Fitch's recent publication, The Evolution of Language.


The first thing you'll notice is that they both feature trees in twilight. Strange enough perhaps, especially when it's not really that obvious what trees have to do with language evolution. However, a further analysis reveals an even greater coincidence:

It appears as though both covers feature the same tree. Indeed, looking up the images revealed that both are property of CORBIS, and both were taken in the Massai Mara Game Reserve, Kenya.

From now on, Language Evolution and Acacia trees will be forever linked because of a chance occurance. There are even rumours of a trip to visit the 'Language Evolution Tree'.

And so structure emerges from chaos ...

Evolutionary Linguists announce arrival of Skynet

In a recent book chapter, Loreto, Baronchelli & Puglisi (2010) summarise mathematical models of language games. They minimally define a system which will allow agents to reach shared linguistic categories which describe continuous stimuli (e.g. colour). New findings include a demonstration that the number of linguistic categories will stay low, regardless of the resolution of the agents' 'eyes'. However, I was particularly struck by the last paragraph, which appears to be ominously more 60s science-fiction than the mathematical jargon:

"Finally it is important to mention that in the last few years a potentially very interesting experimental platform appeared: the World Wide Web. Though only a few years old, the growth of the Web and its effect on the society have been astonishing ..."

What is this wondrous thing called the World Wide Web? Why has nobody told me about it?

Yes, as far as I can tell, this was published last month. Then, the really spooky Skynet stuff:

"Innovation has widened the possibilities for communication. Social media like blogs, wikis, and social bookmarking tools allow the immediacy of conversation, with unprecedented levels of communication speed and community size. In this perspective the web is acquiring the status of a platform for social computing, able to coordinate and exploit the cognitive abilities of the users for a given task."

We are enslaved by this terrible self-organising system! But what task is it trying to complete? Surely nothing good. It seems that Evolutionary Linguists just can't help imagining dystopian futures.


Vittorio Loreto;, Andrea Baronchelli;, & Andrea Puglisi (2009). Mathematical Modeling of Language Games Evolution of Communication and Language in Embodied Agents, 263-281

Language Structure and Social Structure

Last week saw the publication of Lupyan & Dale (2010) (also discussed here). It's an analysis of languages from the World Atlas of Langauge Structures (WALS, quite fun to play around with), showing that languages spoken by more people tend to be less morphologically complex (fewer cases, fewer inflections, more of a tenancy to express things using separate words rather than with morphology). It's hypothesised that this is because a greater, more dispersed population will contain more second-language learners, therefore the language will tend to change to be easier to learn by adult learners, who seem to find morphology more difficult to learn than child learners.

The number of speakers in a language certainly varies a great deal. Lupyan & Dale emphasise this by pointing out that the median number of speakers in a language is 7,000 while the mean is 828,000. The number of second-language learners is also not a trivial influence in modern times. For example, only about 30% of English speakers are native speakers. Even more extreme is Malay with only 15% of speakers learning it as a first language.

In comparison, Siberian Yupik Eskimo has essentially no non-native speakers (from supporting data). In such communities, there is a greater pressure for the language to be learnable by children, therefore it retains morphological complexity.

The changes introduced by adults are likely to be acquired by children. Thus, Lupyan and Dale talk about exoteric and esoteric languages. The changes introduced by adults are likely to be acquired by children. Although the paper also takes into consideration the fact that children do not always learn from their parents:

It has been argued that there is no automatic transmission of the “mother tongue” from parents to offspring (47). For example, in a survey of 188 individuals in Senegal who listed Bambara as their native language, Bambara was the father’s native language in 16%, the mother’s in 19%, the native language of both parents in 26%, and the native language of neither parent in 39% (47).

Interestingly (for me at least), Welsh is pointed out in the regression graph and seems to fit the pattern - it has relatively few speakers and a high complexity score (calculated by "summing the number of features for which each language relies on lexical versus morphological coding and subtracting the total from zero"). While Welsh speakers were dispersed throughout Britain in the 8th century (and a Welsh speaking colony remains in Argentina), there are probably more child speakers than adult speakers currently. However, contact with English has introduced a bias for lexical forms over morphological forms (at least on my schoolyard - consider "Rydw i'n ysgrifennu" vs. "Ysgrifennaf"). Coupled with a pressure from language conservation groups to make Welsh more accessible to learners (see my post here), Welsh may indeed become less morphologically complicated.

Also interesting is the comparison with language data from Ethnologue, noting that the WALS seems to under-represent sub-Saharan Arfican languages (here). It appears that the Ethnologue is still the best source of data on population sizes, even though it appears not to be good enough to calculate the number of bilingual speakers (here).

One question is to what extent this phenomenon of exocentrism is a modern one. Large social groups and extensive travel only really took off in the last thousand years, so it's not clear if these dynamics can be scaled backwards in time to aid study of the evolution of language. It's been argued that human evolution allowed larger group sizes in comparison to chimpanzees (Isbell & Young, 1996), did this force communication to become more structured? Was language primarily lexically-based and used by adults?

Perhaps, however, it could be applied to animals - do songbirds living in larger, migratory flocks have simpler song morphology than those in smaller, more localised ones? The complexity of the domesticated finch is certainly more complex than its wild descendants (Honda & Okanoya, 1999). Could simply reducing the number of 'speakers' be an influence?

It's a fascinating article, and has already stimulated a lot of debate.



Lupyan G, Dale R (2010). Language Structure Is Partly Determined by Social Structure PLoS ONE, 5 (1) : 10.1371/journal.pone.0008559

Honda, E., & Okanoya, K. (1999). Acoustical and Syntactical Comparisons between Songs of the White-backed Munia (Lonchura striata) and Its Domesticated Strain, the Bengalese Finch (Lonchura striata var. domestica) Zoological Science, 16 (2), 319-326 DOI: 10.2108/zsj.16.319

Isbell, L. & Young (1996). The evolution of bipedalism in hominids and reduced group size in chimpanzees: alternative responses to decreasing resource availability Journal of Human Evolution, 30 (5), 389-397 DOI: 10.1006/jhev.1996.0034

Cymraeg Clir

A scheme at Bangor University offers translation services in Welsh. But this one doesn't translate into English. Instead your documents come back in Cymraeg Clir (Clear Welsh). Ok, so it's more of an editorial service really, but there are some interesting features in its guidelines (The document is titled "Ysgrifennu'n Glir" (Writing Clearly), even though it recommends using 'sgrifenny' instead of 'ysgrifennu'):

Move the subject closer to the start of the sentence by using active verbs, rather than passive. This is general advice for any language, but there are Welsh-specific ones too:

Use verbs instead of nouns. This is slightly confusing at first glance, but refers to the use of phrases like 'tree management' rather than the simpler 'managing trees'. The incursion of such nouns is put down to the influence of English.

Use the 'long form' of the verb (uninflected) to avoid tense ambiguity.

Interestingly, the use of acronyms is discouraged, partly for ease of reading but also because acronyms can be misleading, or even difficult to construct, due to initial consonant mutation (the changing of the first consonant in a word in some syntactic contexts, e.g. 'cat' in welsh is cath, but 'his cat' is ei gath). However, established acronyms and English acronyms along with full Welsh translations can be used. Furthermore, acronyms themselves cannot be mutated.

Mutation is also to be avoided in bullet points by inserting e.g. 'the following' before them (all contiguous nouns after a mutation context are elligable for mutation). Interestingly, foreign place names should not be mutated either.

Users are reminded that mutation applies by agreement, not just in immediate contexts, so in the following sentence, since 'neges' is a feminine noun, the verb 'teipio' (to type) must be mutated using 'aspirate' mutation:

Dyma’r neges y mae’n rhaid i mi ei theipio
Here the message it is must to me it type
This is the message that I must type

Whenever rules like this need to be explained in editorial documents, I wonder whether Welsh mutation has just gone too far.

Contracting the verb 'is' (e.g. 'the building's insured') is recommended. However, the Welsh 'yn' (is) can also mean 'within', in which case it should not be contracted. Instead of explaining the secondary meaning in Welsh, however, the guideline uses English:

"heblaw pan fo’r ‘yn’ yn golygu in yn y Saesneg"
(exept when the 'in' means in in English)

Never use the ampersand (&) when writing in Welsh! This makes sense, since it's a contraction of a Latin phrase anyway.

There is also advice on punctuation, including the use of hyphens to separate double letters, avoiding the possibility of triple and quadruple ls (e.g. alllifo).

Ghost in the Shell Tattoos

Following my obsession with Masamune Shirow's Ghost in the Shell, here are some GITS-inspired tattoos. I also found one of the laughing man icon, but couldn't link it. Surprisingly, I couldn't find any individual eleven tattoos, but here's someone who's planning to get one.

By Stanio

By Kryoide

By tetsuonabiki

By Scott White

A Tachikoma from Thursday Morning


By Melissa


By Craig B.


By rkampschoer


By Neon-Stiches


By Todo

From Riokou

By Romantist-Egoist

From M1K3Y

Evolutionary linguists employed by Aliens

As I slowly weave my way deeper into the underworld of Evolutionary Linguistics, I can't help thinking there's an alien conspiracy going on ...

Walking around the basements of Edinburgh Universtiy, signs can be seen inviting you to 'Participate in an Alien Language Experiment'. Indeed, Kirby, Cornish and Smith (2008) tested whether humans could "learn an alien language". In a recent Horizon program, Prof. Kirby can be seen extending his encounters of the third kind in what appears to be a bombed-out Victorian Prison - clearly a secret base for some dark organistation:



Gary Lupyan, recently seen giving a talk in Edinburgh, has run experiments which test whether humans can tell the difference between "approachable" and "dangerous aliens.

Further, in Ichinco, Frank and Saxe (2009)'s study, "participants were told that they had been kidnapped by aliens who were trying to teach them their language through two episodes of alien television". I was under the impression no signals had been received from other planets, never mind double-features.

It also emerged that Evolutionary Linguist Gareth Roberts, had been asking participants to imagine themselves competing for resources on an alien planet.

From this, I can only deduce one thing: Aliens have been employing evolutionary linguists to decide which humans to kidnap and bring back to their planet as slaves. So, the next time you find yourself asked to imagine that you're learning how to mine heavy metals on a distant planet or quizzed about whether you would be likely to oppose oppressors with green tentacles, feign incompetence.

Summary: Grassman and Tomasello (2010)

Grassman and Tomasello recently published a study in reaction to Jaswal & Hansen (2006). Jaswal & Hansen found that children ignore pragmatic cues (pointing and looking at objects) in favour of honouring mutual exclusivity. In the first experiment, 24 children (3-4 y.o) were presented with a novel and a familiar item (see my post on diagrams). They were split into two conditions where the experimenter pointed at the familiar object and either asked 'Can you hand me the X?', where X was a novel noun, or made a neutral request ('Can you give it to me?').

All children in the latter 'pragmatic baseline' condition selected the familiar object the experimenter was pointing or looking at, while almost all children selected the novel object when presented with a novel noun. In a second experiment, a similar setup was used, but where the experimenter shifted their gaze towards one object instead of pointing and used an arbitrary label which could refer to any of the objects (e.g. 'Can you give me the one I got yesterday?') instead of the neutral . Here, children presented with a novel label chose the familiar item significantly less often than those in the pragmatic baseline.

The children were aware of the pragmatic cues: Children in both conditions monitored the experimenter's cue equally often. Furthermore, Jaswal & Hansen report that some children, when asked for a novel noun and presented with a pragmatic cue (looking at the familiar object), checked under the table on the side suggested by the pragmatic cue then selected the novel object anyway.

However, Grassman & Tomasello (2010) question Jaswal and Hansen's conclusions, pointing out that gestural cues are ontogenetically and phylogenetically primary to conventional, linguistic cues.

Furthermore, they argue that Jaswal & Hansen's pragmatic cues were non-ostensive. That is, the experimenter's gaze did not follow the pointing or, in the looking condition, there was no gaze alternation. Thus, there was no cue that the pointing/looking was a communicative act intended for the child. Furthermore, they used 'small' pointing, which is associated with redundant information rather than instead of 'big' pointing which is associated with primary information.

Grassmann & Tomasello extend Jaswal & Hansen's experiments with direct linguistic cues ('Give me the X', where X was a novel noun) instead of indirect ('Give it to me'). Two and four year olds showed no preference for linguistic over non-ostensive cues. However, without pragmatic cues, children chose the the novel object, while with ostensive cues (pointing with whole arm, gaze alternation) children chose the familiar object (see my post on diagrams). These directly oppose Jaswal and Hansen's results.

In a second experiment, children were presented with a familiar and novel object and requested a the familiar object verbally while pragmatically cuing the novel item. Children chose the novel object, following the pragmatic cue.

In a third experiment, the experimenter presented two familiar objects while verbally requesting one and pragmatically cuing another. Here, there were no preferences for following the pragmatic or lexical cue. 2 year olds exhibited mixed responses across trials, while 4 year olds were consistent, but with half following the pragmatic cue and half following the lexical cue.

Grassmann & Tomasello place this data in the context of the child's conflict resolution and lexical entries. Children try to integrate cues from different domains into one coherent communicative intention. It is suggested that it may be harder to modify lexical entries for familiar words without a clear reason than to link novel words to familiar objects.

Jaswal VK, & Hansen MB (2006). Learning words: children disregard some pragmatic information that conflicts with mutual exclusivity. Developmental science, 9 (2), 158-65 PMID: 16472316

Grassmann, S., & Tomasello, M. (2010). Young children follow pointing over words in interpreting acts of reference Developmental Science, 13 (1), 252-263 DOI: 10.1111/j.1467-7687.2009.00871.x

Mutual Exlcusivity Diagrams

Lately, I've been reading about experiments on Mutual Exclusivity. I've found it useful to draw up some diagrams to illustrate each. I've posted some of them below. Each diagram represents entities in the Singal, Concept and Environment domains. Circles represent entities (signals, concepts or actual objects). Lines between circles represent a link between them. For instance the word "Dog" would be linked with a concept "Dog" which could be linked to an actual instantiation of a dog in the environment (as long as the observer had learnt these). Dotted circles indicate a novel entity which the observer has not seen before. Dashed lines indicate links trained by the experimenter.

Au & Glusman (1990):
Byers-Heinlein & Werker (2009):

Jaswal & Hansen (2006):


Grassman & Tomasello (2010) - a reaction to Jaswal & Hansen above:

Healey & Scarabela (2009):


Au TK, & Glusman M (1990). The principle of mutual exclusivity in word learning: to honor or not to honor? Child development, 61 (5), 1474-90 PMID: 2245739

Byers-Heinlein, K., & Werker, J. (2009). Monolingual, bilingual, trilingual: infants' language experience influences the development of a word-learning heuristic Developmental Science, 12 (5), 815-823 DOI: 10.1111/j.1467-7687.2009.00902.x

Jaswal VK, & Hansen MB (2006). Learning words: children disregard some pragmatic information that conflicts with mutual exclusivity. Developmental science, 9 (2), 158-65 PMID: 16472316

Grassmann, S., & Tomasello, M. (2010). Young children follow pointing over words in interpreting acts of reference Developmental Science, 13 (1), 252-263 DOI: 10.1111/j.1467-7687.2009.00871.x

Healey, E. and Scarabela, B. (2009). Are children willing to accept two labels for one object? Proceedings of the Child Language Seminar. University of Reading.

Pun of the Day #6

Trying to get past J with plate of chips

J: I'm afraid to get past me there's a chip tax.
Me: A potatoll?

Synaptic Gasp

A while ago, I posted my own stencil inspired by Ghost in the shell. Here's Synaptic Gasp by ocean.flynn which has similar themes:


ocean.flynn says:

The synaptic cleft in the human brain reminds me of the gap between the hand of God and Adam in Michaelangelo’s visualization of Creation. My mind is stuck on the image of the gap. That’s the leap of faith between that which we can know and that which is beyond our capacity to know.

Inferential Models of Bilingualism

de Boer (2000)'s models showed that shared representations in a communicative system could be achieved given certain principles. Steels & Belpaeme (2005) extended this framework to the domain of colour, which is a bit more intuitive to understand: A speaker and a hearer are placed in an environment full of coloured tiles. A speaker chooses a colour and names it (with the name only the speaker knows), e.g. 'blurby'. The hearer hears the colour name, and tries to match it with a colour tile it can see. It then points to the colour tile it thinks the speaker meant, and the speaker points to the colour tile it actually meant. The hearer then adjusts its internal representations to fit what it has learned. This involves creating a new 'node' placed in a colour map at the location of the colour tile, labelled 'blurby'. From now on, the hearer can recognise any colour close to this as a 'blurby' colour. Over many exchanges like this, individuals come to share the same names for the same colour tiles.

However, no studies have looked at the additional problems produced by bilingualism. What happens when there are two languages being used? Not only can there be two labels for each type of colour tile, but the boundaries between lables can be different (for example, Welsh is one of the many languges that only has one word for the range of colours extending from green to blue).

In fact, no models even allow the possibility of having two labels for the same percept (labels are selected by calculating distances of label foci in a perceptual space). Additionally, most models implicitly assume a mutual exclusivity bias - if agents experience a new label, they assume it refers to a division of the perceptual space that they don't posses and adjust their perceptual space accordingly. However, agents could not necessarily assume this if they were learning two languages.

The solution is not as simple as creating two separate colour maps for each language, since all children have the potential to become bilingual there is no clear default. Also, being bilingual may affect the way you decide which object is being referred to. Studies into the Mutual Exclusivity bias have found that bilinguals are more likely to accept two labels for a single object, especially when they can tell the labels come from different languages (Au & Glusman, 1990; Healey & Scarabela, 2009; Byers-Heinlein & Werker, 2009).

Andrew Smith has investigated the problem of indeterminacy with computational models (Smith, 2005,2005b). The model is based on Steels' (1996) model of perceptually grounded meaning creation. Agents learn labels from each other which refer to perceptions of a simulated environment. Because agents do not have access to each other's internal representations, they need some way of overcoming indeterminacy when faced with a label they do not understand. Smith (2005) suggests that this can be achieved by relying on individual-level representation constraints (perceptual biases) or inferential tactics. One of these inferential tactics is Mutual Exclusivity. Smith models mutual exclusivity by having the agents exclude from consideration all objects for which it already understands a word when perceiving a new word.

The challenge, in light of more recent findings concerning bilinguals, is to define a model which favours mutual exclusivity within languages, but violates it between languages. It's not obvious how this could be achieved, nor how to test it. From previous studies, it's clear that stable 'bilingualism' is unlikely to emerge in such models, so perhaps two systems need to be developed in segregation and then integrated at a test stage.

First, however, there is the problem of detecting which language a word belongs to, or realising that more than one language is being spoken at all. Smith's agents perceive the world through a number of perceptual channels. Although some sort of entropy model could distinguish between two sets of words, it's unclear why this would be useful except for deciding which language was being spoken. In other words, it would only aid the fitting of the model without really explaining how it works.

One way of aiding the learning of labels in bilingual contexts would be to increase the number of perceptual channels to encode the communication context. For instance, an analogue perception of the signal to model language membership cues from prosody (although this still has the same problem as the entropy solution above). Alternatively, some channels could encode a perception of the speaker themselves. Each agent would have perceivable features, analogous to facial features perhaps, which could be used to distinguish which agent was speaking.

The channels could also be extended to include the encoding of a pragmatic cue, enabling a modeling of the experiments in Au & Glusman (1990) and Healey & Skarabela (2009).

Perhaps, however, these models already achieve such phenomena. The models begin with an inherently bilingual society - each agent begins with a minimal, random vocabulary and conceptual divisions. One question is what factors lead to a single, shared system. Is it an inevitable outcome in a cultural system with limits on resources and exposure to stimuli? Or does a shared system emerge because of biases implicit in the model which disfavour bilingualism? There appears to be a gap in the literature anyway, possibly even a PhD-shaped one.

Au TK, & Glusman M (1990). The principle of mutual exclusivity in word learning: to honor or not to honor? Child development, 61 (5), 1474-90 PMID: 2245739
Byers-Heinlein K, & Werker JF (2009). Monolingual, bilingual, trilingual: infants' language experience influences the development of a word-learning heuristic. Developmental science, 12 (5), 815-23 PMID: 19702772
DEBOER, B. (2000). Self-organization in vowel systems Journal of Phonetics, 28 (4), 441-465 DOI: 10.1006/jpho.2000.0125
Grassmann S, Stracke M, & Tomasello M (2009). Two-year-olds exclude novel objects as potential referents of novel words based on pragmatics. Cognition, 112 (3), 488-93 PMID: 19616205
Healey, E. and Scarabela, B. (2009). Are children willing to accept two labels for one object? Proceedings of the Child Language Seminar. University of Reading.
Smith, A. (2005). The Inferential Transmission of Language Adaptive Behavior, 13 (4), 311-324 DOI: 10.1177/105971230501300402
Smith, A.D.M. (2005). Mutual exclusivity: communicative success despite conceptual divergence Language Origins: Perspectives on Evolution

Sitting in the stand

I noticed today a bilingual sign on Cardiff city FC's new stadium, reading "eisteddle Canton stand" (word order on welsh and English is usefully opposite so the sign can be read in welsh as "eisteddle canton" or in English as "canton stand"). What's interesting is that "eisteddle" means "sitting place", the opposite to "stand". As someone pointed out, though, "stand" comes from "grandstand", refering to the building rather than the action. Still, it's interesting that coexisting languages can evolve to label the same thing with labels that oppose each other in other contexts.