Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
The study of evolution and language provides a unique opportunity for carefully examining ba- sic and important questions about evolution, ...
Typology: Slides
1 / 248
This page cannot be seen from the preview
Don't miss anything!
Edward Stabler January 11, 2007
The study of evolution and language provides a unique opportunity for carefully examining basic questions about evolution, language, and the kinds of explanations available for sources of order in physical, biological, cognitive and cultural domains.
The study of evolution and language provides a unique opportunity for carefully examining ba- sic and important questions about evolution, language, and the kinds of explanations available for sources of order in physical, biological, cognitive and cultural domains. Human languages provide a kind of mirror on human thought, and so we want to under- stand the forces that have shaped the structures we see there. Evolution provides a source of structure at two levels. First, the human organism has evolved, with linguistic abilities of certain kinds, by genetic transmission and natural selection. And second, each particular lan- guage is a cultural artifact, transmitted by learning and selected by various cultural and natural forces. In each case, we can ask: what aspects of language structure can be explained by evo- lutionary forces? And what other forces are shaping human languages? These are fundamental questions that every thinking person is likely to be curious about, and so it is no surprise that there is a wealth of popular and scientific work addressing them. Readings will be drawn from classic and contemporary research, supplemented with lecture notes each week.
i
Spring 2006 Syllabus i
A similar shift can be seen in some reactions to the success of the “neo-Darwinian synthesis” in evolutionary biology. A long-standing and still essential component of evolutionary study focuses on the analysis of branching phylogeny (see Figure 1 on page 3): what emerged when, and why; what has been the role of selection and of drift in the history of life? But Gould, Kauffman and others in the 1970’s and 1980’s drew attention to some new kinds of questions: do common properties emerge on different branches of the phylogenetic tree, properties whose emergence must be due to some requirements of form that are independent of selection? There are, and understanding them is key to accounting for the emergence of complex “wholes.” The account must go beyond the standard neo-Darwinian “myriad mutations, selection sifting” to recognize limitations of selection and the importance of other sources of order. This step brings earlier views like D’Arcy Wentworth Thompson’s laws of form back into balance with Darwin’s. This balance is essential for a proper appreciation of the question of how human language and other complex morphology and complex abilities could emerge in the forms we find.
Considering the clearest cases of evolutionary development first, the basic mechanisms of evolution will be studied. We will look briefly at some well-studied examples, including the clear and disastrous case of HIV evolution. This “retrovirus,” with no DNA but only only two identical strands of RNA, is quite different from the large organisms that we are more familiar with, and because of its extremely rapid evolution, AZT and related treatments that succeeded in the short term have all failed in the long term. Such cases of extremely rapid evolution also provide examples of the emergence of dominant variants, “quasi-species.” We then look briefly at some of the basic principles of evolutionary theory, using some examples to illustrate the roles of sources of order other than selection.
With this background, we turn to prominent ideas about the evolution of learning, noticing how selection can achieve a certain balance between rigidity and adaptation in organisms. This sets the stage for a rather careful look at human linguistic ability, which is rigid in some respects and plastic in others. We briefly survey first some of the distinctive features of human languages (features which, for the most part, spoken and signed and written languages all share). Finally, we will be in a position to really understand why the experts have conflicting views about the roles of natural selection, “exaptation,” and laws of form and function in shaping human language.
As a last exercise, we turn to the study of how particular languages, particular cultural artifacts change over time. The tools for studying evolution can be applied to questions in historical linguistics. This is a relatively new field, but one that is booming with the advent of relevant computational methods for simulation, revealing the fundamental interplay between organismic plasticity and cultural transmission.
Figure 1: branching phylogeny calculated from genetic distances (Sogin and Patterson 1992)
Darwin
Darwin sets out his basic ideas out in a clear and summary form in the last chapter of his Origin of Species. We can identify the following basic postulates on which his analysis is based:
We can notice some important differences from Lamarck. First, while Lamarck thought that the apparent adaptation of organisms to their ecological niches resulted from a history of traits acquired by the practice of one’s ancestors, Darwin makes no such assumption. There is undirected variation, and there is selection. These factors alone are held to be responsible for the “creativity” that one seems to see in the adaptations of organisms. Also notice that Lamarck’s second law is not replaced by any other idea about how variation is introduced into a species; it seems to be just provided by nature. Darwin had the view simply that all structures vary, and selection acts on the diversity nature provides.
What evidence is offered in support of these postulates, and the view that natural selection is a basic force behind the diversity of life?
Analogy with breeding and horticulture. This analogy is the main idea in The Origin of Species. The situation for all organisms is like breeding, except first , the selective force is not a human breeder, but the complex of natural forces that determines which organisms will survive to reproduce, and second , the period of time over which natural selection has acted far exceeds human history. Since domestication has given us animals fitting our various needs to such an extent, it is no surprise we see even more exquisite adaptations in the fit between organisms in the wild and their habitat, their “ecological niche.” Dar- win says (§14) “What limit can be put to this power, acting during long ages and rigidly scrutinizing the whole constitution, structure and habits of each creature – favoring the good and rejecting the bad? I can see no limit to this power, in slowly and beautifully adapting each form to the most complex relations of life.”
Fossil records. In some popular accounts of evolution, it is suggested that fossils were Dar- win’s main evidence, but this is very far from the truth. The main evidence comes from the the 3 obvious axioms listed above, and the analogy with breeding and horticulture which shows how successive incremental changes can produce dramatic changes. The fossil evidence actually presents serious difficulties for Darwin’s view that these mech- anisms explain the enormous variation and adaptation that so impressed him, so he considers the problem at length. He suggests that missing intermediate forms and sud- den appearances in the fossil record are plausibly attributed to the imperfection of the fossil record (§9), but nevertheless when fossils are present, adjacent strata tend to dif- fer minimally, while differences between the organisms become larger as one considers strata that are far apart.
Distinct species on islands. Darwin notes that islands frequently have species peculiar to them. “Thus in the Galapagos Islands nearly every land-bird, but only two out of the
eleven marine birds, are peculiar; and it is obvious that marine birds could arrive at these islands more easily than land-birds.” (§12) There are many correlations between geography and distribution of organisms: similar organisms tend to be geographically close to each other, even when the geographically close areas have very different cli- mates. “In considering the distribution of organic beings over the face of the globe, the first great fact which strikes us is, that neither the similarity nor the dissimilarity of the inhabitants of various regions can be accounted for by their climatal and other physical conditions.” (§11)
Vestigial organs. While the “perfection” of adaptation may result from selection over long periods of time, Darwin also observes that apparent “imperfections” may also have an explanation, as ancestral adaptations that are no longer used. “On the view of each organic being and each separate organ having been specially created, how utterly inex- plicable it is that parts, like the teeth in the embryonic calf or like the shrivelled wings under the soldered wing-covers of some beetles, should thus so frequently bear the plain stamp of inutility! Nature may be said to have taken pains to reveal, by rudimen- tary organs and by homologous structures, her scheme of modification, which it seems that we wilfully will not understand.” (§14) Gould calls this the “panda principle” after the peculiar panda’s thumb, and sometimes the “orchid principle” because of the many contrivances Darwin noticed in the petals of orchids.
Homologous organs. The similarities among even very different species calls for some expla- nation: the explanation is that at least many of the similar organs are homologous, that is, they are inherited from common ancestors. Darwin says in §13, “What can be more curious than that the hand of a man, formed for grasping, that of a mole for digging, the leg of the horse, the paddle of the porpoise, and the wing of the bat should all be con- structed on the same pattern, and should include the same bones, in the same relative positions.”
Embryonic similarities. “The points of structure, in which the embryos of widely different animals of the same class resemble each other, often have no direct relation to their conditions of existence. We cannot, for instance, suppose that in the embryos of the vertebrata the peculiar loop-like course of the arteries near the branchial slits are related to similar conditions, in the young mammal which is nourished in the womb of its mother, in the egg of the bird which is hatched in a nest, and in the spawn of a frog under water…As the embryonic state of each species and group of species partially shows us the structure of their less modified ancient progenitors, we can clearly see why ancient and extinct forms of life should resemble the embryos of their descendants, our existing species” (§13)
Many puzzles and problems arise for the theory of natural selection. Some are recognized by Darwin and frankly discussed in his work. Others became clear only later. We will have more to say about all these things.
Given rigorous compositional accounts of simple mathematical languages, it did not take much longer to discover how a physical object could be designed to behave according to the formal rules of such a language – this is the idea of a computer. So by 1936, the mathematician Alan Turing showed how a finite machine could (barring memory limitations and untimely breakdowns) compute essentially anything (any “computable function”). In the short span of 70 or 80 years, these ideas not only spawned the computer revolution, but also revolutionized our whole conception of mathematics and many sciences. Linguistics is one of the sciences that has been profoundly influenced by these ideas: we recognize language structure by computing it, deriving it from our knowledge of the grammar of the language.
As we will see, a human language has some basic units, together with some ways for putting these units together. This system of parts and modes of combinations is called the grammar of the language. With a finite grammar, finite beings like humans can handle a language that is essentially unlimited, producing any number of new sentences that will be comprehensible to others who have a relevantly similar grammar. We accordingly regard the grammar as a cognitive structure. It is the system you use to “decode” the language.
In fact, human languages seem to require compositional analysis at a number of levels: speech sounds are composed from basic articulatory features; syllables from sounds; mor- phemes from syllables; words from morphemes; phrases from words. The semantic composi- tionality is perhaps the most intriguing, though. It is no surprise that it captured the imagina- tions of philosophers early in this century (especially Gottlob Frege, Bertrand Russell, Ludwig Wittgenstein). In effect, a sentence is regarded as an abstract kind of picture of reality, with the parts of the sentence meaning, or referring to, parts of the world. We communicate by passing these pictures among ourselves. This perspective was briefly rejected by radically behaviorist approaches to language in the 1950’s, but it is back again in a more sophisticated form.
Another idea about the compositional structure of language is noted by Frege. He observes that certain parts of sentences require another to be “completed.” For example, negation makes no sense by itself. The following sentences are fine:
(4) It is not the case that the cat likes dogs (5) The cat likes dogs
But if we keep just the underlined part, the result is “incomplete,” and would not usually be said by itself.
(6) * It is not the case that
We use the asterisk to indicate that there is something deviant about having the words It is not the case that by themselves. This string of words is incomplete until it is combined with the cat likes dogs or some other sentence. The same goes for and, or, if…then : these do not occur by themselves.
When we look into the structure of the cat likes dogs , we find a similar thing. The subject of the sentence is the cat and the predicate is likes dogs. The predicate seems to be “incomplete” in the same way as it is not the case that. The predicate likes dogs requires a subject to be
complete. We say that this predicate selects the subject. Going one step further, we can see that the verb likes selects a direct object too, since the cat likes is also incomplete.
In these first simple proposals, there are two very important claims:
(7) The structures of sentences are recursive, in the sense that inside a sentence, other sentences can “recur.” This means that there is no longest sentence, and the language is infinite. Given any declarative sentence, you can make a longer one by adding and and another sentence. (8) Certain parts of a sentence require other parts to be present. In the common technical jargon: Certain parts select other parts.
These simple ideas will be important later.
Like the study of language, the study of things like learning, reasoning, and perception has taken a new shape with the advent of generative and computational models. Finally, these models can be clear and predictive, and even mathematical, where prior work had to be informal and more heavily judgement-laden. Still, these objects of study are very complex, and so it has been difficult to pin them down with the kind of generality and specificity that would be most useful in comparing the abilities of different organisms.
When comparing human cognition with non-human animal cognition, we face the great difficulty that we cannot explore what’s going on by asking the animals about it. And of course we need to guard against attributing our own cognitive abilities to organisms that display similar abilities. A funny example of this is mentioned by (Pinker, 2002, p61), based on work by Laura Petitto, a psychologist who trained and studied a well-known chimpanzee named “Nim Chimpsky.” She actually lived with Nim for a year in a large house in New York state, on a Columbia University research project. Nim seemed to imitate many things that he saw Petitto doing, but not in the way a human child would. For example, seeing Petitto washing the dishes, Nim would imitate her motions and enjoy the warm water. Looking more closely though, it turned out that he had no idea of what the activity was for. He would mimic her motions, rubbing dishes with a sponge, but he never got the idea that the object was to make the dish cleaner.
When we are interested in evolutionary connections between behaviors, there is another kind of confusion that we should guard against: similar behaviors in different organisms can sometimes derive from a common ancestor – in this case they are called “homologies” – but they can also arise independently – in which case they are called “homoplasies.” The inference from mere similarity to an evolutionary connection is not generally a safe one.
Consider, for example, what goes on when you walk across the street: you visually per- ceive various familiar objects in spatial relationships of various kinds, and you coordinate the motions of your muscles in order to move your body. Since animals can navigate their envi- ronments, do they have similar and even homologous abilities to perceive and reason about
Similar puzzles about how local intrinsic properties of materials lead to global regularities come to mind when you contemplate the architecture of crystals and snowflakes, the regular shapes of sand dunes, the spectra of various light sources.
Harold Edgerton (1957) “Milk Drop Coronet”
We see similar emergent global regularities in much more complex phenomena. The structure of proteins provides a relevant example, since, as we will discuss in more detail later, heredity is controlled by DNA and RNA determination of protein synthesis.
All living things, from bacteria to humans, contain proteins built from the 20 naturally occuring amino acids, listed here with their (3 letter and 1 letter) standard abbreviations.
amino acid abbreviations amino acid abbreviations Alanine Ala A Cysteine Cys C Aspartic AciD Asp D Glutamic Acid Glu E Phenylalanine Phe F Glycine Gly G Histidine His H Isoleucine Ile I Lysine Lys K Leucine Leu L Methionine Met M AsparagiNe Asn N Proline Pro P Glutamine Gln Q pARginine Arg R Serine Ser S Threonine Thr T Valine Val V Tryptophan Trp W TYrosine Tyr Y
These amino acids combine in sometimes very long sequences. Sequences of length less than 40 or so are often called peptides. Longer sequences are proteins or polypeptides. These proteins control most important cellular processes. For example, hemoglobin, the protein that carries oxygen in the blood, and which also appears in the cells of plants and even bacteria, has a distinctive sequence of amino acids:
The long polypeptide chains twist, coil and fold up in complex ways, yielding shapes that are typically important determinants of their function. The shape of hemoglobin is thought to be something like this:
Another well-known protein is rhodopsin, found in light-sensitive cells in the eyes of verte- brates and invertebrates, and variants of this protein are even found in algae and in some light-sensitive bacteria. When rhodopsin is exposed to light, the surface of the protein be- comes catalytically active, that is, capable of triggering reactions that change ion distributions
computing, mentioned above) was intrigued by the number of petals on daisies (Turing, 1952).
The leftmost picture shows a shasta daisy, which has 21 petals. Ordinary field daisies have 34 petals, You can find daisies with 21, 34, 55 or even 89 petals, but not 4 or 8 or 25 or 36 petals. As in the previous examples, there is a question about how the number of petals is determined. Does each plant cell “know” how many other petals there are? If not, how can any cell know whether to initiate the development of a petal itself? But the daisies provide an extra clue to how this must work, with an extra puzzle: why these particular numbers of petals?
The numbers of daisy petals listed above are all “Fibonacci” numbers. The Fibonacci num- bers are defined this way
f ib( 0 ) = 1 f ib( 1 ) = 1 f ib(n) = f ib(n − 1 ) + f ib(n − 2 ) for all n > 1
Beginning with the third one, each Fibonacci number is the sum of the previous two. This definition is said to be recursive, since the definition of a Fibonacci number uses the no- tion of Fibonacci numbers. The definition is recursive but not circular, because the definition of each number depends only on earlier values. So we can calculate the values of f ib for n=0,1,2,3,4,…to get the following numbers:
1 , 1 , 2 , 3 , 5 , 8 , 13 , 21 , 34 , 55 , 89 , 144 , 233 ,...
Why are the numbers of petals on a daisy Fibonacci numbers?
Lilies, irises, and the trillium have three petals; columbines, buttercups, larkspur, and wild rose have five petals; delphiniums, bloodroot, and cosmos have eight petals; corn marigolds have 13 petals; asters have 21 petals; and daisies have 21, 34, 55, or 89 petals – all Fibonacci numbers. Flowers with other numbers of petals can be found, but the Fibonacci numbers are wierdly common.
see, e.g. http://ccins.camosun.bc.ca/ jbritton/fibslide/ for more examples
The number of spirals out from the center of a sunflower, a pinecone, a pineapple, or a cauliflower is usually a Fibonacci number. And in some plants with stems leaving a branch at various intervals, the rotation around the stem from one branch to the next is given by a
ratio of successive Fibonacci numbers. Phyllotaxis is the name for such patterns in the leaves or petals or branching patterns of plants. These regularities are clearly not imposed by some external controlling force. One recent proposal about them appears in (Douady and Couder, 1996). (a check on the web will reveal many others too!) Clearly something about the numbers of petals is genetically determined, and this may be influenced by natural selection, but there is some other force acting as well. The matter is still not well understood.
In the metazoa , the kingdom of multicelluar animals, we have similar puzzles. One that was noticed by Turing and has been recently studied by (Meinhardt and Gierer,
Similar puzzles arise in more complex organisms, which typically have several axes of ap- proximate symmetry that originate early in embryonic development. Vertebrates like us have at least 4 axes of symmetry in early embryonic development. Studies suggest that in all these organisms small initial asymmetries due to gravity, the exact point of sperm entry, and other things, get “amplified” somehow to form axes of symmetry for the developing embryo. This requires some kind of communication between cells, via some kind of diffusible substances. The basic idea is that cells developing in certain ways emit chemicals that inhibit certain kinds of changes in other cells. An existing hydra head inhibits the development of another one, and in this case the chemical basis of the signal is fairly well understood.
Again, the thing we see in these cases is an overall plan getting realized by local properties of the parts. In this case, the overall plan is rather complex, and some kind of “communication” among the parts is necessary. Clearly this is genetically determined, in part, may be influenced by natural selection, but the collaboration of physical determinants of axis formation is a com- plex matter, in which each cell relies on a complex of interaction of local stimuli.^2
(^2) There is a flurry of discussion about cell differentiation in vertebrates going on just recently (Dudley, Ros, and Tabin, 2002; Sun, Mariani, and Martin, 2002; Tickle and Wolpert, 2002).
