Where Chris Collins enters the fray

Chris sent me this longish response to some of what has appeared in the blog. In the hope of getting him to become a regularish participant in the ongoing discussions I here post his "Response to Norbert." I feel that he let me off lightly, actually. But this said, I think that I can still finds some points to disagree with. I will restrict these to the comments section and hand the floor over to him. Thx Chris.

*****

Response to Norbert

I read with interest Norbert’s recent post on formalization: “Formalization and Falsification in Generative Grammar”. Here I write some preliminary comments on his post.  I have not read other relevant posts in this sprawling blog, which I am only now learning how to navigate. So some of what I say may be redundant. 

For me the quote by Frege in the Begriffsschrift (pg. 6 of the book “Frege and Godel”) indicates what is important when he analogizes the “ideography” (basically first and second order predicate calculus) to a microscope: “But as soon as scientific goals demand great sharpness of resolution, the eye proves to be insufficient. The microscope, on the other hand, is perfectly suited to precisely such goals, but that is just why it is useless for all others.” Similarly, formalization in syntax is a tool that needs to be employed when needed. It not an absolute necessity and there are many ways of going about things (as I discuss below). By citing Frege, I am in no way claiming that we should aim at the same level of formalization that Frege did.

There is an important connection with the ideas of Rob Chametzky (posted by Norbert) in another place on this blog. As we have seen, Rob divides up theorizing into meta-theoretical, theoretical and analytical.  Analytical work, according to Chametzky is: “concerned with investigating the (phenomena of the) domain in question. It deploys and tests concepts and architecture developed in theoretical work, allowing for both understanding of the domain and sharpening of the theoretical concepts.” It is clear that more than 90% of all linguistics work (maybe 99%) is analytical, and that there is a paucity of true theoretical work.

A good example of analytical work would be Chomsky’s “On Wh-Movement”, which is one of the most beautiful and important papers in the field. Chomsky proposes the wh-diagnostics and relentlessly subjects a series of constructions to those diagnostics uncovering many interesting patterns and facts. The consequence that all these various constructions can be reduced to the single rule of “wh-movement” is a huge advanced, allowing one insight into UG. Ultimately, this paper lead to the Move-Alpha framework, and indirectly to Merge (the simplest and most general operation yet).

However, “On Wh-Movement” is what I would call “semi-formal”. It has semi-formal statements of various conditions and principles, and also lots of assumptions are left implicit. As a consequence it has the hallmark property of semi-formal work: there are no theorems and no proofs. Formalization is stating a theory clearly and formally enough that one can establish conclusively (i.e., with a proof) the relations between various aspects of the theory and between claims of the theory and claims of alternative theories.

Certainly, it would have been a waste of time to fully formalize “On Wh-Movement”. It would have expanded the text 10-20 fold at least, and added nothing. This is something that I think Pullum completely missed in his 1989 paper on formalization. The semi-formal nature of syntactic theory, also found in such classics as “Infinite Syntax” by Ross and “On Raising” by Postal, has led to a huge explosion of knowledge that people outside of linguistics/syntax cannot really understand (hence all the lame discussion out there on the internet and Facebook about what the real accomplishments of generative grammar have been), in part because syntacticians are not very good popularizers.

Theoretical work, according to Rob is:  “is concerned with developing and investigating primitives, derived concepts and architecture within a particular domain of inquiry.” There are many good examples of this kind of work in the minimalist literature. I would say Uriagereka’s original work on multi-spell-out qualifies and so does Epstein’s work on c-command, amongst others.

My feeling is that theoretical work (in Chametzky’s sense) is the natural place for formalization in linguistic theory. The reason is that it is possible, using formal assumptions to show clearly the relationship between various concepts, assumptions, operations and principles. For example, it should be possible to show, from formal work, that things like the NTC and Extension condition should really be thought of as theorems proved on the basis of assumptions about UG.  Since NTC and Extension condition are theorems, they can actually be eliminated from UG. And from this, one can wonder if that program can be extended to the full range of what syntacticians normally think about as constraints.

In this, I agree with Norbert who states: “It can lay bare what the conceptual dependencies between our basic concepts are.” Furthermore, as my previous paragraph makes clear, this mode of reasoning is particularly important for pushing the SMT forward. How can we know, with certainty, how some concept/principle/mechanism fits into the SMT? We can formalize and see if we can prove relations between our assumptions about the SMT and the various concepts/principles/mechanisms. Using the ruthless tools of definition, proof and theorem, we can gradually whittle away at UG, until we have the bare essence. I am sure that there are many surprises in store for us. Given the fundamental, abstract and subtle nature of the elements involved, such formalization is probably a necessity, if we want to avoid falling into a muddle of unclear conclusions.

A related reason for formalization (in addition to clearly stating/proving relationships between concepts and assumptions) is that it allows one to clarify murky areas. One of the biggest such areas nowadays is whether syntactic dependencies make use of chains, multi-dominance structures or something else entirely (maybe nothing else). Chomsky’s papers, including his recent ones, make references to chains at many points. But other recent work invokes multi-dominance. What are the differences and relations between these theories and are either of them really necessary? What assumptions about UG does multi-dominance or chains entail? I am afraid that without formalization it will be impossible to answer these questions. I am investigating these questions in my seminar this semester.

These questions about syntactic dependencies interact closely with TransferPF (Spell-Out) and TransferLF, which to my knowledge, have not only not been formalized but not even stated in an explicit manner. Investigating the question of whether multi-dominance, chains or some something else entirely (perhaps nothing else) is needed to model human language syntax will require a concomitant formalization of TransferPF and TransferLF, since these are the functions that make use of the structures formed by Merge.

Minimalist syntax calls for formalization in a way that previous syntactic theories did not. First, the nature of the basic operations is simple enough (e.g., Merge) to make formalization a real possibility. The baroque and varied nature of “transformations” in the “On Wh-Movement” framework and preceding work made the prospect for a full formalization more daunting.

Second, the nature of the concepts involved in minimalism, because of their simplicity and generality (e.g., copies, occurrences), are just too fundamental and subtle and abstract to resolve by talking through them in an informal or semi-formal way. With formalization we can hope to state things in such a way to make clear conceptual and empirical properties of the various proposals, and compare and evaluate them. In fact, I have recently being doing a lot of this with my colleagues, because only recently (by helping to write Collins and Stabler 2012) have I seen what the issues are.

So, in the spirit of Frege, formalization should be a tool for ordinary working syntacticians to clarify their ideas and examine them empirically and conceptually.

Derivation Trees: Syntacticians' Best Friend?

Usually I would have been busy today bringing the excitement of parts of speech and phrase structure rules to a group of surprisingly energetic students --- some of them take the train all the way from NYC, which according to my calculations means that they have to get up around 5:30 in the morning. But at least today they get to sleep in, as the roads and tracks are covered in this weird, cocaine-colored substance that has been pestering us east coast residents for quite a while now and has even found its way south to Georgia. So classes are canceled, and I'm sitting at home with a glass of milk and a habitual desire to talk about syntax. Good thing there's tons of stuff I haven't told you about Minimalist grammars yet, starting with derivation trees.

More on theory (rantish)

This earlier post quoting Rob Chametzky’s typology of linguistic research generated some interesting discussion, much of it needed in my opinion. I would like to spend a paragraph or two ruminating (ranting might be more like it) on the current state of theoretical work as Rob characterizes it, and in particular, on why there appears to be so little of it. Before starting, let me reiterate that concentrating on theory is not intended to impugn the other kinds of research that linguists do. There is a lot of excellent descriptive and analytic work out there, and three cheers for that!  However, as Peggy pointed out in the comments section, theory is not generally accorded much of a hearing unless it comes from Chomsky, and, IMO, even proposals from this quarter are less well received than they once were. Why?

Peggy offers one very plausible hypothesis: that it is “easier to evaluate analytical work,” which “adopts some premises, applies them within a domain, and analyzes the outcome.” How so? Well because “[t]heoretical work involves examination of premises, and it's much more difficult to convince people that their premises are wrong than to convince them that such-and-such data can be analyzed within their (perhaps slightly amended) premises.” Peggy’s observations express Kuhn’s old observation that “normal science” is, well, the norm, and it generally accepts and adapts given theoretical conceptions rather than challenges them. So, in this regard, theory within linguistics is no different from theory anywhere else.[1] 

However, I am not fully convinced of this. Here’s why. Rob notes that theory itself rests on meta-theory and meta-theory concerns itself both with general methodological concerns (simplicity, consistency, relation to other theories etc.) and with domain specific adequacy conditions.  In the domain of linguistics, the first general methodological concerns have been made prominent within recent minimalist theory, the hard part being how to concretize the methodological concerns in the particular setting of linguistics (e.g. when is a proposal “simpler” or “more elegant” or “less redundant” than another?). Rob illustrates domain specific meta-theory with Chomsky’s differentiating theories that are observationally, descriptively and explanatorily adequate. These meta-theoretical desiderata, especially the third, are where theory lives. I believe that the field has sometimes forgotten this.[2] And if it has, then the dearth of theory should be unsurprising. What then are the large meta-theoretical issues that drive theory?

The first one, which traces back to what Chomsky likes to call “the earliest days of Generative Grammar,” is Plato’s problem (PP). The second, is of more recent vintage, and has been dubbed “Darwin’s Problem” (DP).[3] A theory attains explanatory adequacy (EA) when it can deduce the attested Gs in combination with a specification of the PLD. A theory can be EA+ (‘+’ = ‘beyond’) if the principles the EA theory postulates are ones that did (or at least, plausibly could have) arisen in humans. The PP, DP duo raise theoretical questions all by themselves for they pull in opposite directions; PP feeling comfortable with a richer more linguistically specific FL while DP happier with a poorer less linguistically specific FL. Reconciling this tension is a worthy theoretical project all by itself.

Note that both PP and DP are based on two big, and IMO, hardly contestable facts: viz. (1) that any human can acquire any language in a pretty short time and in pretty much the same way regardless of the language at issue when exposed to PLD of that language, and (2) that human language capacity emerged at some time in the recentish past from ancestors that were not language endowed the way we are. These big facts are (two of) the fixed points of our linguistic meta-theory, and in terms to which theory should be addressed.  This meta-theoretical background places demands both on proposed analyses concerning the structure of particular Gs and on the structure of FL/UG. And it is precisely these demands that allow for the evaluation of proposals somewhat independently of whether they are analytically (in Rob’s sense) sound. In other words, aside from specific familiar linguistic data (e.g. that ‘flying planes can be dangerous’ is ambiguous) that we use to evaluate a given proposal, there is also the question of whether a given proposal can be argued to be acquirable/evolvable. Respect for theory starts with taking these meta-theoretical demands seriously. IMO, our sensitivity to these concerns is currently inappropriately low.

Why do I say this? Here’s some anecdotal evidence for this judgment.

First, I think that many practitioners of the syntactic arts misperceive what the object of inquiry is. If asked: “what does linguistics study?” many will answer: “language.” But language is not the object of study, at least for generative linguists. The faculty of language (FL) is. FL in combination with other cognitive faculties leads to language behavior, utterances, perceptions, plays, movies, etc.  But these products are not the primary object of inquiry despite the fact that studying language behavior, both in the wild and in more artificial settings (e.g. acceptability judgments), has been a good place to find data that bears on the structure of FL. This noted, the goal of generative grammar has never been to describe or regiment language (in fact, many generativists, me included, do not believe that languages are natural kinds and so not appropriated targets of study) but to describe the fine structure of FL.  Now here’s the kicker: if one thinks that the target of inquiry is language, then the theoretical considerations that PP and DP lead to will not seem particularly germane to the enterprise. To address PP and DP we need to advert to the structure of FL and this involves considerations that go beyond covering the data that linguists primarily rely on to make their analytical arguments. Thus, if language replaces FL as the research topic then PP and DP won’t loom so large with the consequence that theory will seem pointless and, thus, not surprisingly, it’s pursuit will be undervalued.

I believe that this shift from FL to language as the cynosure of linguistic inquiry has gotten greater of late. Here’s some anecdotal evidence. There once was a time when the first intro chapter of virtually every thesis in syntax began with a discussion of the logical problem of language acquisition (aka PP) and ended with a concluding chapter considering what the technically meaty chapters 2-5 had implied about UG and Plato’s problem. One might argue that this was mere window dressing and that the formulations and discussions were very pro forma. To a degree, I would agree with this. However, the required discussion (even if cursory) pointed to a (tacit) recognition that the details in the middle were in service of the larger questions driving the field and this served to legitimate these questions and the theory that lives on them.

Nowadays, any similar discussion is hard to find. Indeed, I would go further, the very idea that one’s analytics deserve even cursory consideration in terms of the more encompassing framework concerns is considered sort of quaint. There’s lots of concern of how syntax interfaces with semantics or phonology, lot’s of worries about how structures proposed in language A compare to those in B. But there is relatively little overt worry about PP or DP.

Here’s a question for my senior colleagues: How many times have you asked in a public venue (e.g. at a thesis defense or at a talk), or even over beer, how some proposal you’ve been talking about (with such and such principles and this and that parameters) could be acquired? How many of you in teaching about grammatical variation stop and concentrate on how some rather subtle difference/parameter one is interested in (e.g. the (purported) difference between English and Romance wrt extraction out of weak WH islands) could have been acquired/set? I agree that this is not the only kind of question worth asking and I agree that an analysis might be valuable even in the absence of an answer to this kind of question, but in my recent experience, we act as if this really doesn’t matter at all, which is why such questions are never raised. Indeed, I suspect that many believe that such questions are either BS or are more properly addressed to our psycho-ling colleagues or both (and yes this does suggest a certain kind of unattractive attitude not uncommon to syntacticians).

I would add that in my experience linguists tend to be hostile to theoretical innovation. This is manifest in two ways.

First, we really don’t like having multiple routes to the same conclusion. In other fields, it is considered interesting to reach the same end in two different ways. So there are myriad proofs of the Pythagorean theorem, and all are considered to be of interest. Why? Why would a novel proof still be publishable (and published)? Because it is not only interesting that a certain fact is true (viz. the square of the hypotenuse...) but it is equally (maybe more) interesting how different concepts link together to demonstrate this.[4] Linking concepts together is what theory is all about and the reluctance of linguists to prize this kind of thing betrays a lack of interest in theoretical work.

Second, the field has a severe “historical bias.” What I mean by this is that we demand that later proposals surpass in empirical coverage earlier proposals in order to get a hearing. But why? Why should a newcomer be required to do better than a senior citizen? In fact, let’s go one step further, why shouldn’t a newcomer be given some empirical slack?[5] After all, most of the proposals we prize have been augmented over time to increase their empirical range, so why demand of a new proposal that it cover all the ground of the venerable ancestor and more? Isn't this just a way of making it impossible for new ideas to breathe? And doesn’t this attitude indicate that what we really care about is that the data points be covered rather than how they are covered? And doesn’t this reflect an instrumental conception of theory?

So, in sum, not only do we often act as if having two ways of thinking about a problem is intellectually abhorrent, we often act as if theoretical novelty is (or should be) a punishable offense, novel theory being acceptable only if it brings in its train wider empirical coverage. 

So, that’s why I think that linguists don’t really prize theory, and that’s too bad. It’s unfortunate because it reflects the fact that we have turned away from the foundational questions of the discipline, from the big facts and questions that, IMO, are the problems of deepest interest. Theory is not the only kind of inquiry worth doing, but it has its place and we should once again recognize this. How? Here’s an easy first step: next time you hear a talk or read a paper, ask yourself how the proposal put forward bears on the structure of FL and what kind of light it sheds on PP and/or DP, our two great meta-theoretical questions.

---

[1] Interestingly, for this to take place so readily there must be an assumption that what data are relevant to a proposal is easy to determine without committing theoretical hostages. I am not sure that this is always the case. So a chunk of the controversy surrounding the movement theory of control (something that I am relatively familiar with) hangs on whether certain observations (e.g. partial control) are reflected in syntactic representations or not.  It is not clear that this kind of dispute is a purely data dispute however.

[2] Alex C makes a similar point in the comments section, though he embroiders it in ways I would not.

[3] Actually both were big questions posed at the start of the Generative enterprise. However, as a matter of fact, PP was central to the discussion since (at least) Aspects while DP became important only with the advent of the Minimalist Program. There are good reasons for this (viz. that DP was not really worth discussing until we had some candidate principles of UG). However, right now, both PP and DP are important meta-theoretical framework questions.

[4] Indeed, the route to the conclusion is often more interesting than the conclusion itself. I recall the following (not verbatim) comment from a mathematician after the four-color problem was solved via a computer crunching through all the possibilities: “I guess the problem was not as interesting as we supposed.”

[5] A point that Greg Kobele defends in his thesis.

What real evolution is like

From Bob Berwick: a link to an article indicating just how complicated evolution is. Just-so stories not only fail to deliver the goods, they create an impression at odds with what we now know to be a very complex set of interactions whose fine structure we appear not to have much of a handle on. Yet another reason to be cautious when we tell tall evolutionary tales and criticize proposals on "but how did it evolve?" grounds.

Science automated at last

Funny post from Andrew Gelman (here). Science without scientists. I've always thought that main problem with word processing programs is that they required you to type in whole thoughts. Why not just give it the gist and let the program fill in the missing steps. We are now one step closer. And we eliminate human error and malfeasance. Be the first on your block to get one!

Falsification

The topic is getting some play in the blogosphere due in part to the post on this matter by Sean Carroll (the physicist at Cal Tech, not the eve-devo biologist) on the Edge question of the year (see here and find his contribution on page 202-3). At any rate, there has been some interesting discussion from people in other areas that you might be interested in looking at (here, here, and here). The discussions give subtle takes on Popper's falsificationist dicta, all recognizing that the (at least common) understanding of Popper is way off. Pigliucci notes the problems that the Duhem-Quine thesis raise, i.e. the observation that theories are never tested directly but always rely on many auxiliary assumptions that can serve to insulate a thesis from direct empirical refutation. The discussion is interesting and may serve to temper simple minded resort to the falsificationist gambit.

One thing, however, that seems not to have been mentioned is that quite often we can find evidence for claims that are not falsifiable: for example, the germ theory of disease postulated germs as agents of disease transmission. It's not clear that existential statements are falsifiable, viz. the absence of attested black swans does not imply that black swans don't exist. We make many such existential claims and then go out and look for instances, e.g. germs, black holes, Higg's particles, locality conditions, etc.  In other words, we often actively look for confirmatory evidence for our claims rather than looking to refute them. In fact, I might go further, in the early exploratory phase, finding evidence for one's views is actually more important that finding evidence that could refute them. Good ideas start out life as very fragile. There is always tons of apparent evidence that "refutes" them. Indeed, the more daring the proposal, the more likely it appears to be false (and perhaps the more likely that it is false). So, what does one do? Look for places where it works! And this is a smart thing to do.  It is always fair to ask someone why their proposal matters. One way that it can matter is that it does something interesting, i.e. solves a puzzle, explains recalcitrant data, predicts a new kind of object etc.  Evidence in favor of a proposal is what allows one to make demands on a rational listener. So, falsification as a strategy is useful, but only after a proposal has gained admittance as a live possibility and admission to this august group is paid for in verifications.

At any rate, the discussion is pretty good and there are links to other matters that might matter. It's always nice to see how others flounder with these large methodological concerns. Enjoy.

Being Edgy

For light entertainment, I have just read answers to the Edge question of the year: “What scientific idea is ready for retirement?” Edge.org (here) is the fashionable online “salon” (think French Belle Époque/Early 20^th century) where the illuminati, literati, cognoscenti and scientific elite take on the big issues of the day impresarioed by John Brockman, the academic world’s favorite Rumpelstiltskin; a spinner of dry academic research into popular science gold (and by ‘gold’ I mean $$$$). At any rate, reading the page-lengthish comments has been quite entertaining and I recommend the pieces to you as a way to unwind after a hard day toiling over admission and job files. The answers weigh in at 213 pages of print. Here are a few papers that got me going.

Not surprisingly, two of those that raised my blood pressure were written about language. One is by Benjamin Bergen (BB) (20-1). He is a cog sci prof at UCSD and his proposal for an idea worth retiring is “Universal Grammar.” When I first read this I was really pissed. But I confess that after reading what he took UG to be, I could understand why BB wants it retired.

BB understand UG as claiming two things: (i) that there are “core commonalities across languages” and (ii) that such exist as a matter of “genetic endowment.”  He reports that “field linguists” have discovered that “are much more diverse than originally thought” (who this ‘we’ are is a bit mystifying. Not even a rabid Chomskyan like me has ever doubted that the surface diversity among languages is rather extensive). In particular, not all languages have “nouns and verbs” and not all “embed propositions in others.” In other words, it seems that field linguists (you can see the long shadow of Mr D. Everett here, more anon) have been busy demonstrating something that has been common knowledge for a long long time (and still the common view): that the surface linguistic forms we find across natural language are very diverse and that this diversity of surface forms indicates that there are few surface manifest universals out there. Oddly, BB is happy to concede that “perhaps the most general computational principles are part of our innate language-specific human endowment” but “this won’t reveal much about how language develops in children.”

There is lots to quibble about here: (i) most importantly that this Greenbergian gloss on “universal grammar” is not how people like me (and more importantly, Chomsky) understand UG, (ii) that BB seems not to have read any of the work on Greenberg style Universals common in the current literature (think Cinque hierarchy), (iii) that if UG is correct then this changes the learning problem, (iv) that “inferring the meaning of words” exploits emerging syntactic knowledge that itself piggy backs on the innate computational principles of UG (e.g. Gleitman), etc.  However, putting all of this to one side, I have nothing against giving up BB’s Greenbergian conception of Universal Grammar.

Indeed, I would go further. We should also give up the idea of language as a proper object of inquiry because it is almost certainly not a natural kind. Generative Linguists of the Chomsky stripe should make clear that strictly speaking there is no such thing as English, French, Inuit, etc. and so it is not surprising that these things have no common properties.  BB’s objections are with claims that my team doesn’t make; I (we) don’t suppose that languages universally have certain properties, only that I-languages do. And these properties involve precisely those features that BB seems happy to concede are species specific and biologically given. For my money, I am happy to throw BB’s notion of Universals on the scientific trash heap and would add ‘language’ to the pyre.

As mentioned, BB is clearly channeling Daniel L. Everett (DE) in his comment. DE speaks for himself here (203-205). He wants to dump the idea that “human behavior is guided by highly specific innate knowledge.”[1] You might think from this opening line that the target is once again going to be domain specific principles of UG. But you would be wrong! It seems that what’s got DE riled this time is the very idea of innate characteristics. DE finds any idea of a non-environmental input to development or learning to be illicit. So not only does DE appear to object to domain specific natively given mechanisms, he seems to object to any mental or neural structure at all that is not the result of environmental input. Wow!

I confess, that I found it impossible to make sense of any of this. I can think of no model of development or of learning/acquisition that does not rely on some given biases, however modest, that are required to explain the developmental trajectory.  The argument is not whether such biases are required, but what they look like; hence the discussion concerning domain specificity. But that’s not DE’s position. He wants to dump the distinction between environment and “innate predispositions or instincts” because “we currently have no way of distinguishing” them.  Really? No way? Not even in a particular domain of inquiry?

What are the arguments DE musters? There are three. All piss poor.

First, DE notes that environmental influence is pervasive: “there is never a period in the development individual…when they are not being affected by their environment.” Hence, DE concludes, we cannot currently know what is environmental from what is innately given. Hmm. The problem is complicated, hence unsolvable? The claim that development arises as the joint contribution of input + an initial state and that this means we need to know something about the initial state does not imply that it is easy to decipher what the architecture of the initial state(s) is. DE and I disagree about what the initial state for grammar development is, my UG being very different from his. But no innate principles/biases then no learning/development. So, if you want to understand the latter you need to truck in the former, no matter how hard it is to tease them apart.[2]

Second, it seems that one cannot give an adequate “definition” of ‘innate.’ Every definition has “been shown to be inadequate.” Of course, every definition of everything has been shown to be inadequate. There are no interesting definitions of anything, including ‘bachelor.’ However, there are proposals that are serviceable in different domains and that inquiry aims to refine. That’s what science does. For what I do in syntax, ‘innate’ denotes the given biases/structures required to map environmentally provided PLD into a G.  I have no idea whether these given biases are coded in the genes, are epigenetic, or are handed over to each child by his/her guardian angel. Not that I am not denying that these other questions are interesting and worth investigating. However, for what I do, this is what I mean by innate. Indeed, I suspect that this is what it more or less always means: what needs to be given so that adventitious input can be generalized in the attested ways. Data do not generalize themselves. The principles of generalization must come from somewhere. We call the place they come from the native or instinctual. And though it is an interesting question to figure out how such native information is delivered to the infant, delivered somehow it must be, for without it development/learning/acquisition is impossible.

Third, DE asserts that one cannot propose that some character is innate without “some evolutionary account of how it might have gotten there.” If this is the case, then most of biology and physics might as well stop right now. This view is just nuts! It’s a version of the old show stopper: you don’t know anything until you know everything, which, if true, means that we might as well stop doing anything at all. Let’s for the sake of argument assume that knowing the evolutionary history of a trait is necessary for understanding how a thing works (btw, I don’t believe this: we can know a lot about how something (e.g. wings, bee dances) works without knowing much about how it developed). Even were this the case, it’s simply false that one cannot know about the mechanics of a system without knowing anything at all about how it arose. We know a whole lot about gravity and still don’t know how it “arose.” But, this position is not only false in practice it is methodologically sterile as it endorses the all or nothing view of inquiry, and this, I suspect is why DE proposes it. What DE really wants (surprise, surprise) is to end Chomsky style work in linguistics. He reaches for any argument to stop it. The fact that what he says verges on the methodologically incoherent matters little. This is war, and as in love, for DE, it seems, all things are fair. Read this piece and weep.

As antidote to DE (and Gopnik) it is worth reading Oliver Scott Curry’s contribution (38-9) on Associationism.[3] He writes that associationism is “hollow- a misleading redescription of the very phenomenon that is in need of explanation.” Right on! Curry makes the obvious, yet correct, point that absent a given mechanism that allows one to divide input into the relevant and irrelevant there is no way to use input. Using input requires “prior theory.” A modest point, but given how hard it is to wean people from their empiricist predilections, always a useful on to make.

There are other entries that will infuriate, but I will leave their debunking as an exercise for the reader. For the interested, take a look at N.J. Enfield’s contribution (47-8) heroically defending the view that there is more to “language” than competence.

I should add that the immediately linguistically relevant articles are a small subset of the Edge pieces. Maybe it’s a sign that what current linguists do is not highly prized that there is not a single piece in he lot by anyone I would consider doing serious linguistics. It’s a clear sign that what we do is no longer considered relevant to wider intellectual concerns, at least buy the “Edgy.” This was not always so. Chomskyan linguistics, after all, was once the leading edge (sic!) of the “cognitive revolution.” We really need to do something about this. Maybe I will post on this later. Any suggestions for raising our profile would be welcome.

This said, there are lots of interesting papers in the collection: on the use of stats (75-77, and 176-7), minds and brains (208-9), mysterianism (7-8), big data (24-5 and 176-7), replication (189-90), the scientific method (147-8), science funding (118-9), science vs technology (211-12), unification (88-90), simplicity (168-9, 180-1), elegance (93-4), falsifiability (202-3), the current (very animated and heated) fight over current high theory in physics (every article by the many physicists), among others. The entries are short, and often provocative and entertaining. So, if you are looking for bathroom reading, I cannot recommend this highly enough.

---

[1] Note that DE’s explanandum is “behavior.” But, this is the wrong target for explanation. Steve Pinker’s very nice piece (190-192) puts it very well so let me quote:

More than half a century after the cognitive revolution, people still ask whether a behavior is genetically of environmentally determined. Yet neither genes nor the environment can control the muscles directly. The cause of behavior is the brain. While it is sensible to ask how emotions, motives or learning mechanisms have been influenced by the genes, it makes no sense to ask this of behavior itself.

[2] Allison Gopnik (172-3) has a similarly confusing Edge comment. She too seems to think that the fact that there is a lot of interaction between environmental input and intial state endowments implies that the whole notion of an initial state is misconceived. IMO, her “argument” is little better than DE’s.

[3] See Andy Clark’s piece on I/O models (147) as well.