Tuesday, October 18, 2016

Right sizing ling papers

I have a question: what’s the “natural” size of a publishable linguistics paper? I ask because after indulging in a reading binge of papers I had agreed to look at for various reasons, it seems that 50 is the assumed magic number. And this number, IMO, is too high.  If it really takes 50 pages for you to make your point, then either you are having trouble locating the point that you want to make, or you are trying to make too many of them in a single paper. Why care?

I care about this for two reasons. First I think that the size of the “natural” paper is a fair indicator of the theoretical sophistication of a field. Second, I believe that if the “natural” size is, say, 50 pages, then 50 pages will be the benchmark of a “serious” paper and people will aim to produce 50 page papers even if this means taking a 20 page idea and blowing it up to 50 pages. And we all know where this leads. To bloated papers that make it harder than it should be (and given the explosion of new (and excellent) research, it’s already harder than it used to be) to stay current with the new ideas in the field. Let me expand on these two points just a bit.

There are several kinds of linguistics papers. The ones that I am talking would be classified as in theoretical linguistics, specifically syntax. The aim of such a paper is to make a theoretical point. Data and argument are marshaled in service of making this point. Now, in a field with well-developed theory, this can be usually done economically. Why? Because the theoretical question/point of interest can be crisply stated and identified. Thus, the data and arguments of interest can be efficiently deployed wrt this identified theoretical question/point. The less theoretically firm the discipline the harder it is to do this well and the longer (more pages) it takes to identify the relevant point etc.  This is what I mean by saying that the size of the “natural” paper can be taken as a (rough) indicator of how theoretically successful a field is. In the “real” sciences, only review papers go on for 50 pages. Most are under 10 and many are less than that (it is called “Phys Rev Letters” for a reason). In the “real” sciences, one does not extensively review earlier results. One cites them, takes what is needed and moves on. Put another way, in “real” sciences one builds on earlier results, one does not rehearse them and re-litigate them. They are there to be built on and your contribution is one more brick in a pretty well specified wall of interlocking assumptions, principles and empirical results.

This is less true in theoretical syntax. Most likely it is because practitioners do not agree as widely about the theoretical results in syntax than people in physics agree about the results there. But, I suspect, that there is another reason as well. In many of the real sciences, papers don’t locally aim for truth (of course, every scientific endeavor globally does). Here’s what I mean.

Many theoretical papers are explorations of what you get by combining ideas in a certain way. The point of interest is that some combinations lead to interesting empirical, theoretical or conceptual consequences. The hope is that these consequences are also true (evaluated over a longer run), but the immediate assumption of many papers is that the assumptions are (or look) true enough (or are interesting enough even if recognizably false) to explore even if there are (acknowledged) problems with them. My impression is that this is not the accepted practice in syntax. Here if you start with assumptions that have “problems” (in syntax, usually, (apparent) empirical difficulties) then it is thought illegitimate to use these assumptions or further explore their consequences. And this has two baleful influences in paper writing: it creates an incentive to fudge one’s assumptions and/or creates a requirement to (re)defend them. In either case, we get pressure to bloat.

A detour: I have never really understood why exploring problematic assumptions (PA) is so regularly dismissed.[1] Actually, I do understand. It is a reflex of theoretical syntax’s general anti-theoretical stance. IMO, theory is that activity that explores how assumptions connect to lead to interesting consequences. That’s what theoretical exploration is. If done correctly, it leads to a modicum of explanation.

This activity is different from how theory is often described in the syntax literature. There it is (often) characterized as a way of “capturing” data. On this view, the data are unruly and wild and need to be corralled and tamed. Theory is that instrument used to pen it in. But if your aim is to “capture” the data, then capturing some, while loosing others is not a win. This is why problematic assumptions (PA) are non grata. Empirically leaky PAs are not interesting precisely because they are leaky. Note, then, that the difference between “capturing” and “explaining” is critical. Leaky PAs might be explanatorily rich even if empirically problematic. Explanation and data coverage are two different dimensions of evaluation. The aim, of course, is to get to those accounts that both explain and are empirically justified. The goal of “capture” blurs these two dimensions. It is also, IMO, very counterproductive. Here’s why.

Say that one takes a PA and finds that it leads to a nice result, be it empirical or theoretical or conceptual. Then shouldn’t this be seen as an argument for PA regardless of its other problems? And shouldn’t this also be an argument that the antecedent problems the PA suffers from might possibly be apparent rather than real? All we really can (and should) do as theorists is explore the consequences of sets of assumptions. One hopes that over time the consequences as a whole favor one set over others. Hence, there is nothing methodologically inapposite in assuming some PA if it fits the bill. In fact, it is a virtue theoretically speaking for it allows us to more fully explore that idea and see if we can understand why even if false it seems to be doing useful work.

Let’s now turn to the second more pragmatic point. There has been an explosion of research in syntax. It used to be possible to keep up with, by reading, everything. I don’t believe that this is still possible. However, it would make it easier to stay tuned to the important issues if papers were more succinct. I think I’ve said this on FOL before (though I can’t recall where), but I have often found it to be the case that a short form version of a later published paper (say a NELs or WCCFL version) is more useful than the longer more elaborated descendant.[2] Why? Because the longer version is generally more “careful,” and not always in a good way. By this I mean that there are replies to reviewers that require elaboration but that often obscure the main idea. Not always, but often enough.

So as not to end on too grumpy a note, let me suggest the following template for syntax papers. It answers three questions: What’s the problem? Why is it interesting? How to solve it?

The first section should be short and to the point. A paper that cannot identify a crisp problem is one that should likely be rewritten.

The second section should also be short, but it is important. Not all problems are equally interesting. It’s the job of a paper to indicate why the reader should care. In linguistics this means identifying how the results bear on the structure of FL/UG. What light does your question, if answered, hope to shed on the central question of modern GG, the fine structure of FL.

The last section is the meat, generally. Only tell the reader enough to understand the explanation to the question being offered. For a theory paper, raw data should be offered but the discussion should proceed by discussing the structures that these data imply. GGers truck in grammars, which truck in rules and structures and derivations. A theory paper that is not careful and explicit about these is not written correctly. Many paeprs in very good journals take great care to get the morphological diacritics right in the glosses but often eschew providing explicit derivations and phrase markers that exhibit the purported theoretical point. For GG, God is not in the data points, but in the derivations etc. that these data points are in service of illuminating.

Let me go a bit over the top here. IMO, journals would do well to stop publishing most data, reserving this for available methods addenda available online. The raw data is important, and the exposition should rely on it and make it available but the exposition should advert to it not present it. This is now standard practice in journals like Science and there is no reason why it should not be standard practice in ling journals too. It would immediately cut down the size of most articles by at least a third (try this for a typical NLLT paper for example).

Only after the paper has offered its novelties should one compare what’s been offered to other approaches in the field. I agree that this is suggestion should not be elevated to a hard and fast rule. Sometimes a proposal is usefully advanced by demonstrating the shortcomings in others that it will repair. However, more often than not comparisons of old and new are hard to make without some advanced glimpse of the new. In my experience, comparison is most useful after the fact.

Delaying comparison will also have another positive feature, I believe. A proposal might be interesting even if it does no better than earlier approaches. I suspect that we upfront “problems” with extant hypotheses because it is considered illicit to offer an alternative unless the current favorite is shown to be in some way defective. There is a founder prejudice operative that requires that the reigning champion not be discomfited unless proven to be inferior. But this is false. It is useful to know that there are many routes to a common conclusion (see here for discussion). It is often even useful to have an alternative that does less well.

So, What, Why How with a 15-20 page limit, with the hopes of lowering this to 10-15. If that were to happen I would feel a whole lot guiltier for being so far behind in my reading.

[1] Actually, I do understand. It is a reflex of theoretical syntax’s general anti-theory stance.
[2] This might be showing my age for I think that it is well nigh impossible nowadays to publish a short version of a paper in a NELs or WCCFL proceeding and then an elaborated version in more prestigious journal. If so, take it from me!

Thursday, October 6, 2016

An addendum to the previous post

I want to make two more observations concerning Berlinksi and Uriagereka's (B&U) review of classical case theory.

First, as they emphasize, correctly IMO, what made the Vergnaud theory so interesting as regards explanatory adequacy was that it was not signaled by surface features of DPs in some many languages (e.g. English and Chinese). In other words, it was not WYSIWYG. If it held then it could not be reasonably acquired simply by tracking surface morphology. This is what made it a UG candidate and why it bore on issues of explanatory adequacy. In other words, it was a nice example of PoS thinking: you know it despite no PLD to motivate it, hence it is part of FL/UG. Again, it is the absence of surface reflexes of the principle that made it interesting. As B&U puts it:
Deep down, case is compelling because linguistics has become a part of the Galilean undertaking, a way of explaining what is visible by an appeal to what is not. 
Not being "visible" is the key here.

Second, B&U notes how P&P models were influenced by the work of Monod and Jacob on the operon. Indeed, I would go further: the kind of work that microbiologists were doing were taken to serve as good models of how work on language could proceed and Case theory as Vergnaud envisaged this was a nice example of the thinking. Here's what I mean.

The operon was discovered by research on very simple bacteria and the supposition was made that how it worked there was how it worked everywhere. It's logic extends from bacteria to butterflies, chickens, lions, whales, worms etc. In other words, reasoning based on a very simple organism was taken to illuminate how far different organisms organized their microbiology. And all of this without replicating the work on butterflies, mice, whales etc.  This reasoning as applied to linguistics allows inferences from the intensive study of one language to prima facie apply to all. Indeed, the PoS argument licenses this kind of inference which is why it is such an interesting and powerful form of argument.

Why do I mention this? Because linguists nowadays don't really believe this. Evidence that we don't can be seen in our reactions to critics (like Everett, Evans, Wolfe, Tomasello, etc.). A staple of GG criticism is that it is English centric. The supposition behind this criticism is that one cannot legitimately say anything about FL/UG based on the study of a smattering of languages. To talk about FL/UG responsibly requires studying a broad swath of different languages for only in so doing is one licensed to make universal inferences. We reply to the critics by noting how much current linguistic work is typological and cross linguistic and that so many people are working on so many different  kinds of languages. But why is this our only retort. Why not say that one can gain terrific insight into FL/UG by studying a single language? Why the requirement that any claim be founded on masses of cross linguistic investigation?

Note that this is exactly what Monod and Jacob did not do. Nor do microbiologists do so today. Microbiologists study a handful of model organisms and from these we infer laws of biology. That is deemed ok in biology but not linguistics. Why? Why do linguists presuppose that only the extensive study of a wide number of different languages will allow insight into FL/UG? It's not the history of the field, so far as I can tell.

B&U shows how classical case theory arose. Similar stories can be told for virtually every other non-trivial theory within linguistics. It arose not via the study of lots of languages but by trying to understand simple facts within a small number in some deep way. This is how bounding theory arose, the ECP, binding and more. So why the presupposition (visible in the replies we give to our critics that we do, really really do, study more than just English) that cross linguistic typological investigations are the only sure way to investigate FL/UG?

I think that I know one answer: we don't really know much about FL/UG. In other words, many linguists will reply that our claims are weak. I don't buy this. But if you do then it is not clear why GGs critics upset you with their claims. Is it that they are saying out loud what you believe but don't think should be shared in polite company?

Wednesday, October 5, 2016

A zesty intro to the logic of explanatory adequacy with a nod to JR Vergenaud

David Berlinski and Juan Uriagereka have written a very readable (and amusing) history of abstract case theory (here). It is also very pedagogical for it focuses on how Vergnaud's proposal regarding abstract case enhanced the explanatory power of UG, and it does this by showing how Chomsky-Lasnik filters were a vast improvement over ordered rules with all of their intricacies and how case theory was a big conceptual improvement over filters like *[NP to VP].  It is all very nicely done (and quite funny in a very dry sort of way.

The story ends with the observation that ECM is, well "exceptional" and suggests, coyly, that this raises interesting issues. It does. One of the nicest results of recent minimalist theory, IMO, was Lasnik and Saito's (L&S) regularization of Postal's scope facts wrt ECM subjects in the context of a theory of case that junks government and replaces it with something like the old spec-head configuration. What L&S show is that given such a theory, one that the earliest versions of MP promoted, we would expect a correlation between (abstract) case value and the scope of the case assigned DP.  Postal's data, L&S argued showed exactly that. This was a wonderful paper and one of the first really interesting results of minimalist logic.

As you all know, this result fit ill with the mover to Agree based Probe-Goal conceptions of case licensing (after all, the whole idea of the L&S theory is that the DP had to move to a higher position in order to get case licensed and this movement expanded its scope horizons). Chomsky's more recent ideas concerning labeling might force object movement as well and so reclaim the Postal, though not within the domain of the theory of case.  At any rate, all of this is to indicate that there are further interesting theoretical movements prompted by Vergnaud's original theory even to the present day. And yes, I know that there are some who think that it was entirely on the wrong track but even they should appreciate the Berlinksy-Uriagereka reconstruction.

Form and function; the sources of structure

I just read a fascinating paper and excellent comment thereupon in Nature Neuroscience (thx to Pierre Pica for sending them along) (here and here). The papers make the interesting point that, as I will argue below, illuminate two very different views of what structure is and where it comes from.  The two views have names that should now be familiar to you: Rationalism (R) and Empiricism (E). What is interesting about the two papers discussed below is that they indicate that R and E are contrasting philosophical conceptions that have important empirical consequences for very concrete research. In other words, R and E are philosophical in the best sense, leading to different conceptions with testable empirical (though not Empiricist) consequences. Or, to put this another way, R and E are, broadly speaking, research programs pointing to different conceptions of what structure is and how it arises.[1]

Before getting into this more contentious larger theme, let’s review the basic findings. The main paper is written by the conglomerate of Saygin, Osher, Norton, Youssoufian, Beach, Feather, Gaab, Gabrielli and Kanwisher (Henceforth Saygin et al). The comment is written by Dehaene and Dehaene-Lambertz (DDL). The principle finding is, as the title makes admirably clear, that “connectivity precedes function in the development of the visual word form area.” What’s this mean?

Saygin et al observes that the brain is divided up into different functional regions and that these are “found in approximately the same anatomical location in virtually every normal adult”.[2] The question is how this organization arises: “how does a particular cortical location become earmarked”? (Saygin et al:1250). There are two possibilities: (i) the connectivity follows the function or (ii) the function follows the connectivity. Let’s expand a bit.

(i) is the idea that in virtue of what a region of brain does it wires up with another region of brain because of what it does at roughly the same time. This is roughly the Hebbian idea that regions that fire together wire together (FTWT). So, a region that is sensitive to certain kinds of visual features (e.g. Visual Word From Area (VFWA)) hooks up with an area where “language processing is often found” (DDL:1193) to deliver a system that undergirds reading (coding a dependency between “sounds” and “letters”/”words”). 

(ii) reverses the causal flow. Rather than intrinsic functional properties of the different regions driving their connectivity (via concurrent firing), the extrinsic connectivity patterns of the regions drives their functional differentiation. To coin a phrase: areas that are wired together fire together (WTFT). This is what Saygin et al finds :

This tight relationship between function and connectivity across the cortex suggests a developmental hypothesis: patterns of extrinsic connectivity (or connectivity fingerprints) may arise early in development, instructing subsequent functional devel­opment.

The may is redeemed to a does by following young kids before and after they learn to read. As DDL summarizes it (1193):

To genuinely test the hypothesis that the VWFA owes its specializa­tion to a pre-existing connectivity pattern, it was necessary to measure brain connectivity in children before they learned to read. This is what Saygin et al. now report. They acquired diffusion-weighted images in children around the age of 5 and used them to reconstruct the approximate trajectory of anatomical fiber tracts in their brain. For every voxel in the ven­tral visual cortex, they obtained a signature pro­file of its quantitative connectivity with 81 other brain regions. They then examined whether a machine-learning algorithm could be trained to predict, from this connectivity profile, whether or not a voxel would become selective to written words 3 years later, once the children had become literate. Finally, they tested their algo­rithm on a child whose data had not been used for training. And it worked: prior connectivity predicted subsequent function (my bold, NH). Although many children did not yet have a VWFA at the age of 5, the connections that were already in place could be used to anticipate where the VWFA would appear once they learned to read.

I’ve bolded the conclusion: WTFT and not FTWT. What makes the Saygin et al results particularly interesting is their precision. Saygin et al is able to predict the “precise location of the VWFA” in each kid based on “the connectivity of this region even before the functional specialization for orthography in the VWFA exists” (1254). So voxels that are not sensitive to words and letters before kids learn to read, become so in virtue of prior (non functionally based) connections to language regions.

Some remarks before getting into the philosophical issues.

First, getting to this result requires lots of work, both neuro imaging work and good behavioral work. This paper is a nice model for how the two can be integrated to provide a really big and juicy result.

Second, this appeares in a really fancy journal (Nature Neurosceince) and one can hope that it will help set a standard for good cog-neuro work, work that emphasizes both the cognition and the neuroscience. Saygin et al does a lot of good cog work to show that in non-readers VWFA is not differentially sensitive to letters/words even though it comes to be so sensitive after kids have learned to read.

Third, DDL points out (1192-3) that whatever VWFA is sensitive to it is not simply visual features (i.e. a bias for certain kinds of letter like shapes).  Why not? Because (i) the region is sensitive to letters and not numerals despite letters and numerals being formed using the same basic shapes and (ii) VWFA is located in the same place in blind subjects and non-blind ones so long as the blond ones can read braille or letters converted into “synthetic spatiotemporal sound patterns.” As DDL cooly puts it:

This finding seems to rule out any explanation based on visual features: the so-called ‘visual’ cortex must, in fact, possess abstract properties that make it appropriate to recognize the ‘shapes’ of letters, numbers or other objects regardless of input modality.

So, it appears that what VWFA takes as a “shape” is itself influenced by what the language area would deem shapely. It’s not just two perceptual domains with their independently specifiable features getting in sync, for what even counts as a shape depends on what an area is wired to. VWFA treats a “shape” as letter-like if it tags a “shape” that is languagy.

Ok, finally time for the sermon: at the broadest level E and R differ in their views of where structure comes from and its relation to function.

For Es, function is the causal driver and structure subserves it. Want to understand the properties of language, look at its communicative function. Want to understand animal genomes, look at the evolutionarily successful phenotypic expressions of these genomes. Want to understand brain architecture, look at how regions function in response to external stimuli and apply Hebbian FTWT algorithms.  For Es, structure follows function. Indeed, structure just is a convenient summary of functionally useful episodes. Cognitive structures reflect shaping effects of the environmental inputs of value to the relevant mind. Laws of nature are just summaries of what natural objects “do.” Brain architectures are reflections of how sensory sensitive brain regions wire up when concurrently activated. Structure is a summary of what happens. In short, form follows (useful) function.

Rs beg to differ. Es understand structure as a precondition of function. It doesn’t follow function but precedes it (put Descartes before the functional horse!). Function is causally constrained by form, which is causally prior. For Rs, the laws of nature reflect the underlying structure of an invisible real substrate. Mental organization causally enables various kinds of cognitive activity. Linguistic competence (the structure of FL/UG and the structure of individual Gs) allows for linguistic performances, including communication and language acquisition. Genomic structure channels selection.  In other words, function follows form. The latter is causally prior. Structure  “instructs” (Saygin et al’sterm) subsequent functional development.

For a very long time, the neurosciences have been in an Empiricist grip. Saygin et al provides a strong argument that the E vision has things exactly backwards and that the Hebbian Esih connectionist conception is likely the wrong way of understanding the neural and functional structure of the brain.[3] Brains come with a lot of extrinsic structure and this structure casually determines how it organizes itself functionally. Moreover, at least in the case of the VWFA, Darwinian selection pressures (another kind of functional “cause”) will not explain the underlying connectivity. Why not? Because as DDL notes (1192) alphabets are around 3800 years old and “those times are far too short for Darwinian evolution to have shaped our genome for reading.” That means that Saygin et al’s results will have no “deeper” functional explanations, at least as concerns the VWFA. Nope, it’s functionally inexplicable structure all the very bottom. Connectivity is the causal key. Function follows. Saygin et al speculate that what is right for VWFA will hold for brain organization more generally. Is the speculation correct? Dunno. But being a card carrying R you know where I’d lay my bets.

[1] I develop this theme in an article here.
[2] This seems like a pretty big deal to me and argues against any simple minded view of brain plasticity, I would imagine. Maybe any part the brain can perform any possible computation, but the fact that brains regularly organize themselves in pretty much the same way seems to indicate that this organization is not entirely haphazard and that there is method behind it. So, if it is true that the brain is perfectly plastic (which I really don’t believe) then this suggested that it is not the computational differences responsible for its large scale functional architecture. Saygin et al suggest another causal mechanism.
[3] In this it seems to be reprising the history of immunology which moved from a theory in which the environment instructed the immune system to one in which the structure of the immune system took causal priority. See here for a history.

Sunday, October 2, 2016

More on the irrelevance of Everett's research to UG, and a long critical review of Wolfe's lousy book

Here are two pieces you might find interesting.

First, Chomsky has recently been interviewed by the NYT about the Everett/Wolfe junk we have spent far too much time on (Thx to Chomsky for the quote). Here he is asked about Everett's work.

You have mentioned one paragraph that Wolfe got right in his book...what was in that paragraph? Was it an explanation of your work? Why do you think we're seeing this resurgence of analysis? You must get fairly tired of defending your work?!
It was a paragraph in which he quoted my explanation to him of why his crucial example, the Amazonian language Piraha, is completely irrelevant to his conclusions and claims.  The reason is quite simple.  Whether the alleged facts about the language are correct or not (apparently not), they are about the language, not the faculty of language, while the general principles he thinks are being challenged have to do with the faculty of language, explicitly and unambiguously, including the work he cites, and speakers of this language share the common human faculty of language, of course, as illustrated by their fluency in Portuguese.  So his entire article and book are irrelevant to his claims.  To take an analogy, if some tribe were found in which everyone wears a black patch over one eye, it would have no bearing on the study of binocular vision in the human visual system.  The problems in this work extend far beyond the total irrelevance of his examples to his claims, but I won’t elaborate here.I’ve been defending the legitimacy of this work, extensively and in print, for 60 years.  In earlier years the discussion were with serious philosophers, linguists, cognitive scientists.  I’m sorry to see that the resurgence you mention does not begin to approximate that level, one reason why unlike earlier years, I don’t bother to respond unless asked.
I have italicized the most important point: Note: it does not matter if Everett is right because his claims are irrelevant even if correct. This is the critical point and one that has, sadly, been obscured in most discussions. Not that the point has not been made. It has been. Rather, the point is quickly made and then the falsity of Everett's claims are discussed at length. This leaves the appearance that the empirical issues matter to the big point at hand. How? Because the space dedicated to the arguing for their falsity swamps that dedicated to their irrelevance. People conclude that if it really didn't matter then why spend all that time worrying about whether the claims are true. Chomsky's reply is economical and to the point. My suggestion: if you really want to debate the ins and outs of Piraha do so in a separate very technical paper that makes clear that it has nothing to do with UG.

Here is a second review online of Wolfe's book brought to my attention by Peter Ludlow. It is very amusing. I especially like the last paragraph for it identifies the real scandal in all of this.
I’m not worried about Chomsky, however, no more than I’m worried about Darwin’s position in future histories of science. Chomsky’s position too will be just fine. I do worry about how we will look in those histories, however. Because from where I sit the rampant anti-intellectual responses and the failures to distinguish nonsense from solid science in the attacks on Chomsky’s work look more like harbingers of a new dark age, one that rejects thoughtful scientific probes into human nature and levels charges of a new kind of apostasy– the apostasy of using one’s mind instead of gut instinct. And I suspect that, from the perspective of future intellectual historians, Chomsky’s ability to produce this last great piece of work against the backdrop of our new dark age will make his achievements seem all the more impressive.
The shoddiness of our high brow press is the real scandal, and it is one that deserves much more serious attention.

Wednesday, September 28, 2016

Vox cognoscenti

The thoroughly modern well informed member of the professional classes reads Vox (I do). Not surprisingly, then, Vox has reviewed Wolfe’s book to provide discussion fodder for those that frequently eat out in groups. Unlike the NYT or The Chronicle, Vox found a linguist, John McWhorter (at Columbia) to do the explaining. The review (henceforth JMR) (here) makes three points: (i) that Wolfe really has no idea what he is talking about, (ii) that GGers are arrogant and dismissive and their work (overly) baroque, and (iii) that Everett and other critics might really have a point, though (and here JM is being fair and balanced) so does Chomsky and Everett’s critics.

The biggest problem with JMR lies with its attempts to split the difference between Chomsky and his “critics.” As I’ve noted before (and the reason I put critics in scare quotes), Everett (and Vyvyan Evans who also gets a mention) have no idea what Chomsky has been claiming and so their “critiques” have little to do with what he has actually said. They thus cannot be criticisms of Chomsky’s work and are thus of little value in assessing Chomsky’s claims.

Furthermore, it is clear that these critiques are of interest to the general public and are covered by the high brow media precisely to the extent that they show that Chomsky is wrong about the nature of language, as he understand this. Indeed, this is why part of every title to every piece covering these critiques declares that Chomsky is wrong about Universal Grammar (UG). Nobody outside a small number of linguists really cares about whether Piraha embeds sentences! All the fireworks in the high-brow press are due to what Everett’s findings mean for the Chomsky program, which is precisely nothing for it rests on a simple misunderstanding of Chomsky’s claims (see here and here for recent discussion).

The sociological significance of the expansive coverage of these “critiques” given their shoddiness is another matter. It says a lot about how much our thought leaders want to discredit Chomsky’s non-linguistic views. I would go further: I doubt that our thought leaders care much about the fine structure of the Faculty of Language. But they hope that discrediting Chomsky’s scientific/linguistic project might also serve to discredit his non-linguistic ones. However, given FoL’s remit, I won’t here develop this (pretty obvious) line of speculation. Instead I will lightly review some JMR highlights and make the (by now, I hope) obvious points. 

A. JMR describes the position that Wolfe is attacking (via Everett’s work) as follows:

Wolfe’s topic is Noam Chomsky’s proposal that all humans are born with a sentence structure blueprint programmed in their brains, invariant across the species, and that each language is but a variation upon this "universal grammar" generated by an as-yet unidentified "language organ." In other words, we are born already knowing language. (2-3)

This is a very misleading way of putting Chomsky’s claims about UG. A better analogy is that there is a biologically given recipe for constructing Gs on the basis of PLD. These Gs can (and do differ) significantly. The “blueprint” analogy suggests that all Gs have the same structures, with a tweak here or there (JRM: “few “switches” that flip in the toddler’s brain”). And this suggests that finding a language with a very different G “blueprint” (Piraha say, which JRM (reporting on Everett) writes does not allow for “the ability to nest ideas inside one another” (with an example of multiple sentential embedding as illustration)) would constitute a problem for Chomsky’s FL conception as it would fail to have a key feature of UG (“the absence proves that no universal grammar could exist”). But, as you all know, this is incorrect. It is consistent with Chomsky’s views that the “blueprints” differ. What matters is that the capacity to draw them (i.e. acquire Gs) remains the same. Put more directly, JMR suggests a Greenbergian understanding of Chomsky Universals. And that, is a big no-no![1]

Now, to be honest, I sometimes had trouble distinguishing what JMR is reporting from what JMR is endorsing. However, as the whole point of the non-debate is that what Everett criticizes is at right angles to what Chomsky claims, leaving this fuzzy severely misreports what is going on. Especially when JMR reports that Chomsky “didn’t like this,” thereby suggesting that it was the content of Everett’s claim that Chomsky objected to, rather than the logic behind it. Chomsky’s primary objection was, and still is, that even if Everett is right about Piraha, it has nothing to do with GG claims about whether recursion is built into the structure of FL/UG.

This is the important point about Everett’s research, and it must be highlighted in any informative review. Once this point is firmly and clearly made one can raise secondary issues (very secondary IMO): whether Everett’s specific claims about Piraha are empirically accurate (IMO, likely not). However, this is decidedly a secondary concern if one’s interest is the relevance of Everett’s claims to Chomsky’s claims concerning the structure of FL/UG. JMR fails to make this simple logical point. Hence, whatever its other virtues, it serves to obscure the relevant issues and so to misinform.

            B. JMR writes that the “meat of the debate” revolves around “Chomskyans belief that adaptations have arisen in the brain that serve exclusively to allow speech.” This is contrasted with views that believe that “speech merely piggybacks on equipment that already evolved to allow advanced thought.”

There are some small bones to pick with this description. Thus, the issue is not speech but linguistic knowledge more generally. But let’s put this aside. Is there really a disagreement between Chomsky and his critics about how linguistically specific FoL is? I doubt it. Or, more accurately, if there is such a disagreement Chomsky’s critics have had nothing whatsoever to say about it. Why?

The question is an interesting one and, as you all know, it is the central question animating the Minimalist Program (MP). MP takes as a research question whether FoL is entirely reducible to operations and primitives of general cognition and computation or whether there is a linguistically specific residue despite FoL’s computational operations largely overlapping with general principles of computation and cognition. 

Before addressing how would one go about resolving this debate, let me observe (again) how modest the Chomskyan claim is. It does not say that every part of FoL is linguistically specific. It does not deny that language interacts with other areas of cognition, emotion or culture. It does not assert that every detail of linguistic competence or behavior is insulated from everything else we know and do. Nope. It makes the very modest claim that there is something special about language, something that humans have qua being human and that this is interesting and investigatable.

Of course, over the years linguists have made specific proposals concerning what this something special might be and have identified properties of FoL that don’t look to be easily reducible to other cognitive, computational, emotional or cultural factors. But this is what you would expect if you took the question seriously. And you would expect those that took the opposite view (i.e there is nothing linguistically special about human linguistic facility) to show how to reduce these apparent linguistically sui generis facts to more general facts about cognition, computation or whatever. But you would be wrong. The critics almost never do this. Which suggests, that there really is no serious debate here. Debate would require both sides to address the question. So far as I can tell, critics interpret Chomsky as claiming that culture, general cognition etc. have no impact on any part of language knowledge or use. They then go on to point to cases where this appears to be false. But as Chomsky never denied this, as his claim is far more modest, these observations, like those of Everett’s concerning recursion, are beside the point. To have a debate, there must be some proposition being debated. So far as I can tell, once again this is false in this particular case. Hence no debate.

JMR notes that dealing with the substantive question of the linguistically specificity of FoL requires getting empirically and theoretically quite technical.[2]

…without a drive-by of this rather occult framework, one can’t begin to understand the contours, tone, and current state of the debate Wolfe covers. (8)

Of course, JMR is correct. How could it be otherwise? After all, if one is arguing that the computations are linguistically sui generis then one needs to specify what these are. And, not surprisingly, these investigations can get quite technical. And JMR understands this. However, it also seems to find this offensive. Note the occult. Later on JMR says:

…from one academic generation to the next, this method [standard GG analyses:NH] of parsing language has mission-crept into a strangely complicated business, increasingly unrelated to what either laypeople or intellectuals outside of linguistics would think of as human language. It is truly one of the oddest schools of thought I am familiar with in any discipline; it intrigues me from afar, like giant squid and 12-tone classical music. (10)

Hmm. JMR clearly suggests that things have gotten too technical. A little is ok, but GGers have gone overboard. JMR seems to believe that dealing with the question about FoL’s specific fine structure should be answerable without getting too complex, without leaving the lay person behind, without technical intricacies. Imagine the reaction to a similar kind of remark if applied to any other scientific domain of inquiry. Reminds me of the Emperor’s quip to Mozart in Amadeus: Sorry Mr Mozart, too many notes!

JMR concedes that all of this extra complexity would be fine if only there was evidence for it.

The question is whether there is independent evidence that justifies assuming that speech entails these peculiar mechanisms for which there is no indication in, well, how people talk and think.

And the problem is that this independent evidence does not seem to exist; anyway, outsiders would find it peculiar how very little interest practitioners have in demonstrating such evidence. Rather, they stipulate that syntax should be this way if it is to be "interesting," if it is to be, as the literature has termed it, "robust" or "rich." Yet where does the idea that how we construct sentences must be "robust" or "rich" in the way this school approves of come from? It’s an assumption, not a finding. (13)

This is calumny. If there is one thing that linguists love to do is find empirical consequences of some piece of formal machinery. But, with this summary judgment, JMR joins the Everett/Evans camp and simply asserts that it is too much. There really are too many notes- “Split IP, Merge, phases and something called “little v”” (14). That these proposals come backed by endless empirical justification is hardly mentioned, let alone discussed. Look, I love hatchet jobs, but as JMR notes about Wolfe, even a drive-by heading to this conclusion requires more than assertion.

I suspect that JMR includes this to be able to play both sides of the fence: sure Wolfe knows nothing, but really he is somewhat right. No. He isn’t. Nor is JMR’s suggestion that there is something to Wolfe’s suspicions justified or, IMO, justifiable.

C. Then there is the linguists and their “bile” against anyone “questioning universal grammar” (16). More specifically against Everett.

On a personal note, I did not take any interest in Everett’s findings until I read the New Yorker piece, and then only because of how badly it misrepresented matters. Nor do I believe that anyone else would have noticed it much, but for the public brouhaha. Even then, had the high-brow press not used Everett’s work to denigrate my own, I would have given it a free pass. But this is not what occurred. The claim was repeatedly made that Everett’s work demonstrates that GG is wrong. Efforts to show that this is incorrect have not been greeted nearly as enthusiastically. JMR mischaracterizes the state of play. And in doing so, once again, obscures the issues at hand.

The “GGers (Chomkyans) are vitriolic” trope has become a staple of the “GG/Chomskyan linguistics is dead” meme. Why? There is one obvious reason. It allows Chomsky’s critics to shift debate from the intellectual issues and refocus them on the personal ones (i.e. to move from content to gossip). To argue against the claims GG has made requires understanding them. It also takes a lot of work because there is a lot of this kind of work out there. Saying that GGers are meanies allows one to stake the high ground without doing any of the intellectual or empirical hard work. This is not unlike political coverage one finds in the press: lots of personality gossip, very little policy analysis.

Why is JMR so sensitive? Apparently some students at some conference found what they were hearing uninteresting (18). Though, JMR notes that “most Chomkysans” are not as dismissive (19). Yet he mentions that there does exist “a current of opinion within the Chomskyan syntax orbit that considers most kinds of linguistic inquiry as beside the point” (19). So, some students are bored about anything outside their immediate interests and some linguists are dismissive. And this means what exactly? It means that GGers are vitriolic and dismissive (though most aren’t), or they could be because such dismissiveness is in the air. This is really dumb stuff, not even People Magazine level titter.

D. Towards the end, JMR notes that Everett has not really made his case concerning Piraha (25-6). Indeed, he finds the Nevins, Pesetsky and Rodrigues rebuttal “largely convincing” and believes that it is “quite plausible that Piraha is not as quirky a language as Everett proposed” (26). Here my views and JMR’s coincide. However, to repeat, JMR’s reasonable discussion mainly serves to obscure the main issue . Let me end with this (again).

JMR, like many of the other discussions in the high-brow press frame the relevant debate in terms of whether Everett is right about Piraha embedding. This presupposes that what Everett found out (or not) about Piraha is relevant to Chomsky’s claims. Because the coverage is framed as an empirical debate, when GGers dismiss Everett’s claims they can be described as having acted inappropriately. They have failed to dispassionately consider the empirical evidence, evidence which the coverage regularly reports would undermine central tenets of Chomsky’s theory of FL/UG if accurate. But this framing is wrong. There is no empirical debate because the presupposition is incorrect. What has gotten (some) GGers hot under the collar is this mis-framing. It is one thing to be shown to be wrong. It is quite another to have people debunk views that you have never held and then accuse you of being snippy because you refuse to hold those views. I don’t dismiss Everett’s views because I fear they might be right. I dismiss these views because they are logically irrelevant to the claims I am interested in (specifically whether (some kind of) recursion is a linguistically specific feature of FL) and because every time this is pointed out the critic’s feelings get hurt. That really does boil the blood.

To end: As reviews go, JMR is not the worst. But it is bad enough. And part of what makes it bad is its apparent judiciousness. It follows the standard tropes and frames the issues in the now familiar ways, albeit with a node here and there to Chomsky and GG. But, as noted, that is the problem. It really seems to be hard for many to accept that so much contention in the press can be based on a pun (GUs vs CUs) and that the whole “debate” is intellectually vapid. But that’s the way it is. Let’s hope this is the last round for the time being.

[1] If memory serves, Stephen Jay Gould discussed a similar problem in biology with the notion blueprint. He noted that for a long time genetic inheritance was conceived in blueprint terms. This, Gould argued, led to homoncular theories of genetic information transmission (we each had a smaller version of ourselves deep down that contained the relevant genetic pattern). This made sense, he argued, if we thought in terms of blueprints. Once we shifted to thinking in terms of codes, the homunculus theory disappeared. I have no idea where Gould made this point, but it has interesting parallels in current conceptions of UG as blueprints.
[2] I’ve discussed this issue before and noted how one might go about trying to adjudicate it rationally (see here).

Friday, September 23, 2016

Chomsky was wrong

An article rolled across my inbox a few weeks ago entitled Chomsky was Wrong. Here's a sequence of six short bits on this article.

Yes, Chomsky said this

When I read the first sentence that new research disproved Chomsky's claim that English is easy - that it turns out English is a hard language - I thought it was a parody. To start with, Chomsky said no such thing. But, indeed, it's about a real paper - even if that paper is about English orthography. In SPE (Sound Pattern of English), Chomsky and Halle claimed that English orthography was "near-optimal": that it reflected pretty closely the lexical representations of English words, except where the pronunciation was not predictable.

That's not what you expect to be reading about when you read about Chomsky. For one thing, there's a reason that there's a rumour Chomsky didn't even write any of SPE. The rumour is pretty clearly false, but Chomsky never worked in phonology again, and he certainly didn't write anything close to all of SPE. For another thing, after all the attacks on "Chomsky's Universal Grammar," it's jarring to read a rebuttal of a specific claim.

But here it is. Let's give both credit and blame where credit is due. Chomsky's name's on the book, so he's responsible for what's in it. If it's wrong, then fine. Chomsky was wrong.

The paper in question is sound

The paper in question is by Garrett Nicolai and Greg Kondrak of the University of Alberta, and it's from the 2015 NAACL (North American Association for Computational Linguistics), linked here.

Nicolai and Kondrak have a simple argument. Any spelling system that's isomorphic to the lexical representations of words should also be "morphologically consistent." That is, the spelling of any given morpheme should be the same across different words. Of course: because multi-morphemic words, at least according to SPE, are built by combining the lexical representations of their component morphemes. Regardless of what those lexical representations are, any spelling that reflects them perfectly will have perfect morphological consistency. English spelling, it turns out, doesn't have this property.

As Chomsky and Halle observed, though, there's a reason that this perfect transparency might not hold for a real spelling system: the spelling system might also want to tell you something about the way the word is actually pronounced. In words like deception, pronounced with [p] but morphologically related to deceive, which is pronounced with [v], you can have a morphologically consistent spelling in which the dece- morpheme has a p in deception, or in which it has a in deceive, (or neither), but you can't have both and still be morphologically consistent. And yet, morphological consistency can make the pronunciation pretty opaque. And so Nicolai and Kondrak have a way to evaluate what's driving English spelling's lack of morphological consistency is perhaps a dose of reader-saving surface-pronunciation transparency. It's not.

The paper is nice because it gets around the obvious difficulty in responding to arguments from linguists, which is that they are usually tightly bound to one specific analysis. Here the authors have found a way to legitimately skip this step (reflecting the underlying forms - thus at least being morphologically consistent - except when really necessary to recover the surface pronunciation). It's a nice approach, too. The argument rests on the constructibility a pseudo-orthography for English which maximizes morphological consistency except when it obscures the pronunciation - exactly what you would expect from a "near-optimal" spelling system - a system that turns out to have much higher levels of both morphological consistency and surface-pronunciation transparency than traditional English orthography. I review some of the details of the paper - which isn't my main quarry - at the bottom below for the interested.

Hanlon's razor

Sometimes you see scientific work obviously distorted in the press and you say, I wonder how that happened - I wonder what happened between the interview and the final article that got this piece so off base. No need to wonder here.

A piece about this research (a piece which was actually fine and informative) appeared on the University of Alberta news site (Google cached version) about a year after the paper was published. Presumably, the university PR department came knocking on doors looking for interesting research. The university news piece was then noticed by CBC Edmonton, who did an interview with Greg Kondrak on the morning show and wrote up a digested version online. The author of this digested version evidently decided to spice it up with some flippant jokes and put "Chomsky was wrong" in the headline with a big picture of Chomsky, because people have heard of Chomsky.

The journalist didn't know too much about the topic, clearly. In an earlier version Noam Chomsky was "Norm Chomsky," Morris Halle was "Morris Hale," and the photo caption under Chomsky was missing the important context - about how the original claim was, in the end, an insignificant one - and so one could read, below Chomsky's face, the highschool-newspaper-worthy "This is the face of a man who was wrong." And, predictably, "English spelling" is conflated with "English", leading to the absurd claim that "English is 40 times harder than Spanish."

The awkward qualification that now appears in the figure caption ("This is the face of a man who - in a small segment from a book published in 1968 - was wrong") bears the mark of some angry linguists with pitchforks complaining to the CBC. Personally, I don't know about the pitchforks. Once I realized that the paper was legit, I wasn't able to muster raising a hackle about the CBC article. It doesn't appear to be grinding an axe, just a bad piece of writing. If it weren't for the fact that, in many other quarters, the walls echo with popular press misinformation about generative linguistics which is both damaging and wrong, this article wouldn't even be a remote cause for concern.

It is possible to talk about Chomsky being wrong without trying to sell me the Brooklyn Bridge

The Nicolai and Kondrak paper, and the comments they gave to the university news site, show that you can write something that refutes Chomsky clearly, and in an accurate, informed, and mature way. The content of the paper demonstrates that they know what they're doing, and have thought carefully about what Chomsky and Halle were actually saying. In the discussion, nothing is exaggerated, and no one is claiming to be the winner or exaggerating their position as the great unlocker of things.

Contrast this with Ibbotson and Tomasello's Scientific American article. Discussed by Jeff in a three-part series recently on the blog by Jeff, it purports to disprove Chomsky. It should be possible to write a popular piece summarizing your research program without fleecing the reader, but they don't. When the topic is whether Chomsky is right or wrong, fleece abounds.

Let me just take three egregious instances of simple fact checking in the Scientific American article:
  • "Recently ... cognitive scientists and linguists have abandoned Chomsky’s 'universal grammar' theory in droves"
    • (i) abandoned - as in, previously believed it but now don't - (ii) in droves - as in, there are many who abandoned all together - and (iii) recently. I agree that there are many people who reject Chomsky, and (i) is certainly attested over the last 60 years, but (ii) and (iii), or any conjunction of them, is totally unfounded. It feels like disdain for the idea that one should even have to be saying factually correct things - a Trump-level falsehood.
  • "The new version of the theory, called principles and parameters, replaced a single universal grammar for all the world’s languages with ..."
    • There was never any claim to a single grammar for all languages.
  • "The main response of universal grammarians to such findings [about children getting inversion right in questions with some wh-words but not others] is that children have the competence with grammar but that other factors can impede their performance and thus both hide the true nature of their grammar"
    • I know it's commonplace in describing science wars to make assertions about what your opponent said that are pulled out of nowhere, but that doesn't make it right. I so doubt that the record, if there is one, would show this to be the "main response" to the claim they're referring to, that I'm willing to call this out as just false. Because this response sounds like a possible response to some other claim. It just doesn't fit here. This statement sounds made up.

This article was definitely written by scientists. It contains some perfectly accurate heady thoughts about desirable properties of a scientific theory, a difficult little intellectual maze on the significance of the sentence Him a presidential candidate!?, and it takes its examples not out of noodling a-priori reasoning but actually out of concrete research papers. In principle, scientists are careful and stick to saying things that are justified. And yet, when trying to make the sale, the scientist feels no compunction about just making up convenient facts.

Chomsky and Halle's claim sucks

The statement was overblown to begin with. C&H are really asking for this to be torn down. Morphological-consistency-except-where-predictable is violated in the very examples C&H use to demonstrate the supposed near optimality, such as divine (divinE), related in the same breath to divinity (divin- NO e -ity), which is laxed under the predictable trisyllabic laxing rule.

But the claim can, I think, further, be said to "suck" in a deeper way in the sense that it

  1. is stated, in not all but many instances as it's raised throughout the book, as if the lexical forms given in SPE were known to be correct, not as if they were a hypothesis being put forward
  2. is backed up by spurious and easily defeasible claims, convenient for C&H if they were true - but not true.
Some examples of (2) are the claim on page 49 that "the fundamental principle of orthography is that phonetic variation is not indicated where it is predictable by general rule" - says who? - and the whopper in the footnote on page 184 (which also contains examples of (1)):

Notice, incidentally, how well the problem of representing the sound pattern of English is solved in this case by conventional orthography [NB: by putting silent e at the end of words where the last syllable is, according to the SPE analysis, [+tense], but leaving it off when it's [-tense], while, consistent with the SPE proposal, leaving the vowel symbol the same, in spite of the radical surface differences induced by the vowel shift that applies to [+tense] vowels]. Corresponding to our device of capitalization of a graphic symbol [to mark +tense vowels], conventional orthography places the symbol e after the single consonant following this symbol ([e] being the only vowel which does not appear in final position phonetically ...). In this case, as in other cases, English orthography turns out to be [!] rather close to an optimal system for spelling English. In other words, it turns out to be rather close to the true [!] phonological representation, given the nonlinguistic constraints that must be met by a spelling system, namely, that it utilize a unidimensional linear representation instead of the linguistically appropriate feature representation and that it limit itself essentially to the letters of the Latin alphabet.

Take a minute to think about the last statement. There are many writing systems that use two dimensions, including any writing system that uses the Latin alphabet with diacritics. In most cases, diacritics are used to signify something phonetically similar to a given sound, and, often, the same diacritic is used consistently to mark the same property across multiple segments, much like a phonological feature. Outside the realm of diacritics, Korean writing uses its additional dimension to mark several pretty uncontroversial phonological features. As far as being limited to letters of the Latin alphabet goes - let's assume this means for English, and not really for "spelling systems"  - just as with diacritics, new letters have been invented as variants of old ones, throughout history. My guess is that this has happened fairly often. And, after all - if you really felt you had to insert an existing letter to featurally modify an old one, presumably, you would stick the extra letter next to the one you were modifying, not as a silent letter at the end of the syllable. 

As for making it sound like the theory was proven fact, maybe it's not so surprising. Chomsky, in my reading, seems to hold across time pretty consistently to the implicit rhetorical/epistemological line that it's rarely worth introducing the qualification "if assumption X is correct." Presumably, all perceived "fact" is ultimately provisional anyway - so who could possibly be so foolish as to take any statement claiming to be fact at face value? I don't really know if Chomsky was even the one responsible for leaving out all the instances of "if we're correct, that is" throughout SPE. But it wouldn't surprise me. But Chomsky isn't alone in this - Halle indulges in the same, and, perhaps partly as a result, the simplified logic of 1960s generative phonology to suppose on the basis of patterns observed in a dictionary that one has all the evidence one needs about the generalizations of native speakers is still standard, drawing far too little criticism in phonology. Hedging too much in scientific writing is an ugly disease, but there is such a thing as hedging too little.

In the current environment, where we're being bombarded with high profile bluster about how wrong generative linguistics is, it's worth taking a lesson from SPE in what not to do. Chomsky likes to argue and he's good at it. Which means you never really lose too much when you see him valuing rhetoric over accuracy. He's fun and interesting to read, and the logic of whatever he's saying is worth pursuing further, even if what he's saying is wrong. And you know he knows. But if an experimental paper rolled across my desk to review and it talked about its conclusions in the same way SPE does, only the blood, of the blood, sweat and tears that would go into the writing of my review, would be metaphorical.

Make no mistake - if it comes to a vote between a guy who's playing complicated intellectual games with me and a simple huckster, I won't vote for the huckster. But I won't be very happy. Every cognitive scientist, I was once cautioned, is one part scientist and one part snake oil salesman.

Nicolai and Kondrak is a good paper, and, notably, despite being a pretty good refutation of a claim of Chomsky's, it's a perfectly normal paper, in which the Chomsky and Halle claim is treated as a normal claim - no need for bluster. And the CBC piece about it is a lesson. If you really desire your scientific contribution to be coloured by falsehood and overstatement, you're perfectly safe. You have no need to worry, and there's no need to do it yourself. All you have to do is send it to a journalist.

Some more details of this paper

Here is a graph from Nicolai and Kondrak's paper of the aforementioned measures of morphological consistency ("morphological optimality") and surface-pronunciation transparency ("orthographic perplexity") - closer to the origin is better on both axes:

The blue x on sitting on the y axis, which has 1 for orthographic perplexity (but a relatively paltry 93.9 for morphemic optimality), is simply the IPA transcription of the pronunciation. The blue + sitting on the x axis, which has 100 for morphemic optimality (but a poor 2.51 for orthographic perplexity), is what you would obtain if you simply picked one spelling for each morpheme, and concatenated them as the spelling of morphologically complex words. (The measure of surface-pronunciation transparency is obviously sensitive to how you decide to spell each morpheme, but for the moment that's unimportant.)

Importantly, the blue diamond is standard English orthography ("traditional orthography", or T.O.), sitting at a sub-optimal 96.1 morphemic optimality and 2.32 orthographic perplexity (for comparison, SR and SS, two proposed spelling reforms, are given). On the other hand, Alg, the orange square, is a constructed pseudo-orthography that keeps one spelling for each morpheme except where the pronunciation isn't predictable, in which case as few surface details as possible are inserted, which leads to a much better orthographic perplexity of 1.33, while maintaining a morphemic optimality of 98.9. This shows that there's no obvious excuse for the lack of morphological consistency.

What keeps English orthography from being optimal? If you apply some well-known English spelling rules it's easy to see. If in your calculation of morphological consistency you ignore the removal the final silent e in voice etc which disappears in voicing, the spelling of panic (and other words that can be followed by -ing to violate the consistency of the pronunciation of -ci- as [s]) as panick instead, and the replacement of the y in industry and other similar words with i in industrial), along with a few other obvious changes, then English orthography pops up to 98.9 percent morphemic optimality, the same level as Alg.

Those spelling rules should attract the attention of anyone who's read SPE, as they're tangentially related to the vowel shift rule, the velar softening rule, and the final yod, but the fact is in all these cases the spelling, with respect to the SPE analysis of these words, reflects the lexical representation in the base form and the surface pronunciation in the derived form. Well, sure, which means that these alternations are presumably at least throwing the orthography a bone as far as its pronunciation transparency goes, but you can do way, way, way better. That's the point. English orthography may be better than it could be if it were maximally morphologically consistent, but it doesn't seem to be optimal.

For details of the measures you can have a look at the paper.