Relativized Minimality and similarity based interference

The goal of the Minimalist Program is to reduce the language specificity of FL (i.e. UG) to the absolute minimum, so as to reduce the Logical Problem of Language Evolution (LPLE) to manageable size. The basic premise  is that LPLE is more challenging the more linguistiky FL’s operations and principles are. This almost immediately implies that to solve the LPLE requires radically de-linguistifying the structure of FL.

There are several reasonable concerns regarding this version of LPLE, but I will put these aside here and assume, with Chomsky, that the question, though vague, is well posed and worth investigating (see here and here for some discussion). One obvious strategy for advancing this project is to try and reduce/unify well grounded linguistic principles of UG with those operative in other domains of cognition. Of the extant UG principles ripe for such reconceptualization, the most tempting, IMO and that of many others as we shall see, is Relativized Minimality (RM). What is RM to be unified with? Human/biological memory. In particular, it is tempting to see RM effects as what you get when you shove linguistic objects through the human memory system.[1] That’s the half-baked idea. In what follows I want to discuss whether it can be more fully baked.

First off, why think that this is doable at all? The main reason is the ubiquity of Similarity Based Interference (SBI) effects in the memory literature. Here is a very good accessible review of the relevant state of play by Van Dyke and Johns (VD&J).[2] It seems that human (in fact all biological) memory is content addressable (CA) (i.e. you call the memory in terms of its contents (rather than, say, an index)). Further, the more the contents of specific memories overlap, the more difficult it is to successfully get at them. More particularly, say that one accesses a memory via certain content cues, the more these cues overlap the more they “overload” the retrieval protocol making it harder to successfully get the right one. On the (trivial) assumption that memory will be required to deal with the ubiquitous non-(linearly)-adjacent dependencies found in language we should expect to find restrictions on linguistic dependencies that reflect this memory architecture.[3] VD&J review various experiments that show the effects that distracters can have on retrieving the right target when these distracters “resemble” the interacting expressions.

Given that memory will create SBIs it is natural to think that some kinds of dependencies will be favored over others by this kind of memory architecture. Which? Well ones in which the cues/features relating the dependents are dissimilar from those of the intervening elements. Graphically, (1) represents the relevant configuration. In (1), if non-adjacent X and Y need to be related (say there is a movement or antecedence dependency between the two) then this will be easiest if the cues/features relating them are not also shared by intervening Zish elements.

(1)  …X…Z…Y…

This should look very familiar to any syntactician who has ever heard the name ‘Luigi Rizzi’ (and if you haven’t think of either changing fields or getting into a better grad program). Since Rizzi’s 1990 work, (1), in the guise of RM, is standardly used to explain why WHs cannot move across other WHs (e.g. superiority and Wh-island effects) or heads over heads (the head movement constraint).

IMO, RM is one of the prettiest (because simplest) empirically useful ideas to have ever been proposed in syntax.[4] Moreover, its family resemblance to the kinds of configurations that induce SBI effects is hard to miss.[5] And the lure of relating the two is very tempting, so tempting that resistance is really perverse. So the question becomes, can we understand RM effects as species of SBI effects and thus reflections of facts about memory architecture?

Psychologists have been pursing a similar (though as we shall see, not identical) hunch for quite a while. There is now good evidence that VD&J reviews that encumbering (working) memory with word lists while sentences are being processed differentially affects processing times of non-local dependencies and that the difficulty is proportional to how similar the words in memory are with words in the sentence that need to be related. Thus, for example, if you are asked to keep in memory the triad TABLE-SINK-TRUCK while processing It was the boat that the guy who lived by the sea sailed after two days then you do better at establishing the dependency between boat and sail then if you are asked to parse the same sentence with fix in place of sail. Why? Because all three of the memory list words are fixable, while none are sailable. This fact makes boat harder to retrieve in the second fix sentence than the first sail sentence (p. 198-9).

Syntactic form can also induce interference effects. Thus, the subject advantage inside relative clauses (RC) (i.e. it is easier to parse subject relatives than object relatives, see here) is affected by the kinds of DPs present in the RC.  In particular take (2) and (3). The Subject Advantage is the fact that (2) is easier to parse.

(2)  The banker that praised the barber climbed the mountain 

(3)  The banker that the barber praised climbed the mountain

VD&J note that SBI effects are detectable in such cases as the Subject Advantage can be reduced or eliminated if in place of D-NP nominal like the barber one puts in pronouns, quantified DPs like everyone and/or proper names. The reasoning is that the definite descriptions interfere with one another, while the names, pronouns and quantifiers interfere with D-NPs far less.[6]

VD&J offers many more examples making effectively the same point: that the contents of memory can affect sentential parsing of non-local dependencies and that they do so by making retrieval harder.

So, features matter, both syntactic features and “semantic” ones (and, I would bet other kinds as well).

What are the relevant dimensions of similarity? Well, it appears that many things can disrupt, including grammatical and semantic differences. Thus, the “semantic” suitability of a word on the memorized word list and the syntactic features that differentiate one kind of nominal from another can serve to interfere with establishing the relation of interest. [7]

Friedmann, Belletti and Rizzi (FBR) (here) reports similar results, but this time for acquisition. It appears, for example, that subject relatives are more easily mastered than object relatives, as are subject vs object questions. FBR discusses data from Hebrew. Similar results are reported for Greek by Varlokosta, Nerantzini and Papadopoulou (VNP) here. Moreover, just as in the processing literature, it appears that DPs interfere with one another the more similar they are. Thus, replacing D-NP nominal with relative pronouns and bare whs (i.e. what vs what book) eases/eliminates the problem. As FBR and VNP note, the subject advantage is selective and, in their work, is correlated with the syntactic shapes of the intervening nominal.[8] The more similar they are, the more the problems caused.

So, at first blush, the idea that RM effects and SBI effects are really the same thing looks very promising. Both treat the shared features of the interveners and dependents as the relevant source of “trouble.” However (and you knew that knew was coming, right?) things are likely more complicated. What’s clear is that features do make a difference, including syntactic ones. However, what’s also clear is that not only syntactic shape/features matters. So do many other kinds.

Moreover, it is not clear which similarities cause problems and which don’t. For example, the standard RM model (and the one outlined in FBR) concentrates on cases where the features are identical vs when they overlap vs when they are entirely disjoint. The problem with relative clauses like (3) for example, is that the head of the relative clause and the intervening subject have the exact same syntactic D-NP shape and the reason subbing a pronoun or name or quantifier might be expected to mitigate difficulty is that the subject intervener only share some of their features thereby reducing the minimiality effects. So in the case of RCs the story works as we expect.

The problem is that there are other data to suggest that this version of RM delivers the wrong answers in other kinds of cases. For example, a recent paper by Atkinson, Apple, Rawlins and Omaki (here) (AARO) shows that “the distribution of D-linking amelioration effect [sic] is not consistent with Featural Relativized Minimality’s predictions…” (1). AARO argues that carefully controlled rating methods of the experimental syntax variety show that moving a which-NP over a which-NP in Spec C is better than moving it over a who (i.e. (4) is reliably better than (5)). This is not what is expected given the featural identity in the first case and mere overlap in the second.[9]

(4)  Which athlete did she wonder which coach would recruit

(5)  Which athlete did she wonder who would recruit

IMO, superiority shows much the same thing. So (6) is quite a bit better than (7) to my ear.

(6)  I wonder which book which boy read

(7)  I wonder which book who read

Once again, the simple syntactic version of RM suggests that the opposite should be the case. If this is so, then there is more than just structural similarity involved in RM effects.

This, however, might be a good thing if one’s aim is to treat RM effects as instances of more general SBI effects. We expect many different factors to interact to provide a gradation of effects, with syntactic shape being one factor among many. The AARO data suggests that this might indeed be correct, as are the parallels between the VD&J parsing data and the FBR/VNP acquisition data. So, if AARO is on the right track, it argues in favor of construing RM effects as kinds of SBI effects, and this is what we would expect were RM not a grammatically primitive feature of FL/UG, but the reflection of general memory architecture when applied to linguistic objects. In other words, this need not be a problem, for this is what one would expect if RM were just a species of SBI (and hence traceable to human memory being content addressable).[10]

What is more problematic, perhaps, is settling what “intervention” means. In the memory literature intervention is entirely a matter of temporal order (i.e. if Z is active when X and Y are being considered it counts as an “intervener,” i.e. roughly speaking if Z is temporally between X and Y then Z intervenes), while for RM the general idea is that intervention is stated in terms of c-command (i.e. Z intervenes between X and Y if X c-commands Z and Z c-commands Y) and this has no simple temporal implications. Thus, the notion explored in the memory literature mainly explores a “linear” notion of intervention while RM relies on a structural notion and so it is not clear that RM effects should be assimilated to memory effects.

However, I am not currently sure how big a problem this might be. Here’s what I mean.

First, much of the literature reviewed in VD&J involves English data where linear and hierarchical intervention will be the same. We know that when these two are pulled apart in many cases it is hierarchy that matters (see the discussion of the Yun et. al. paper here. It shows that the Subject Advantage persists in languages where linear order and hierarchical order go in different directions).

Similarly, Brian Dillon (here) and Dave Kush (in his unavailable thesis; ask him) show that hierarchical, not linear, intervention is what’s critical in computing binding relations.

Of course, there are also cases where hierarchy does not rule, and linear intervention seems to matter (e.g. agreement attraction errors and certain NPI licensing illusions are less sensitive to structural information, at least in some online tasks, then hierarchical restrictions suggest they should be).[11]

So both notions of proximity seem to play a role in language processing. I don’t know whether both have an effect in acquisition (but see note 11). So, does this mean that we cannot unify RM effects with SBI effects for they apply in different kinds of configurations?  Maybe not.  Here’s why.

Memory effects arise from two sources: the structure of memory (e.g. is it content addressable, rates of decay, number of buffers, RAM etc.) and the data structures that memory works on. It is thus possible that when memory manipulates syntactic structures that it will measure intervention hierarchically because linguistic objects are hierarchically structured. In other words, if phrase markers are bereft of linear order information (as say a set theoretic understanding of phrase markers entails) then when memory deals with these it will not be able to use linear notions to manipulate them because such objects have no linear structure. In these cases, when task demands use memory to calculate the properties of phrase markers, then RM effects is what we expect to see: SBIs with c-command determining intervention.  Of course, sentences when used have more than hierarchical structure and it is reasonable to suppose that this too will affect how linguistic items are used.[12] However, this does not prevent thinking of RM as a memory effect defined over PM-like data structures. And there is every reason to hope that this is in fact correct for if it is then we can treat RM as a special case of a more general cognitive fact about us; that we have content addressable memories that are subject to SBI effects. In other words, we can reduce the linguistic specificity of FL.

Maybe an analogy will help here. In a wonderful book, What the hands reveal about the brain (here), Poizner, Bellugi and Klima (PBK) describe the two ways that ASL speakers with brains damage describe spatial layouts. As you all know, space has both a grammatical and a physical sense for an ASL speaker. What PBK note is that when this space is used grammatically then it functions differently than when it is used physically. When used physically, stroke patients with spatial deficits show all the characteristic problems that regular right hemisphere stroke patients show (e.g. they use only half the space). However, when signing in the space (i.e. using the space linguistically) then this particular spatial deficit goes away and patients no longer neglect half the signing space. In other words, depending on how the space is being used, linguistically or physically, determines what deficits are observed.

Say something analogous is true with memory; when used in computing grammatical properties, intervention is hierarchical. When used otherwise, linear/temporal structure may arise.  Thus what counts as an intervener will depend what properties memory is being used to determine. If something like this makes sense (or can be made to make sense) then unifying RM effects as SBI effects with both related to how human memory works looks possible.[13]

Enough. This post is both too long and too rambling. Let me end with a provocative statement. Many at Athens felt that Darwin’s Problem (the LPLE) is too vague to be useful. Some conceded that it might have poetic charm, that is was a kind of inspirational ditty. Few (none?) thought that it could support a research program. As I’ve said many times before, I think that this is wrong, or at least too hasty. The obvious program that LPLE (aka Darwin’s Problem) supports is a reductive/unificational one. To solve LPLE requires showing that most of the principles operative in FL are non-linguistically specific. This means showing how they could be reflections of something more cognitively (or computationally or physically) general. RM seems ripe for such a reanalysis in more general terms. However, successfully showing that RM is a special case of SBI which is grounded in how human memory operates will take a lot of work, and it might fail. However, the papers I’ve cited above outline how redeeming this hunch might proceed. Can it work? Who knows. Should it work? Yup, the LPLE/DP hangs on it.[14]

---

[1] The first person I heard making this connection explicitly is Ivan Ortega Santos. He did this in his 896 paper at UMD in about 2007 (a version published here). It appears that others were barking up a similar tree somewhat earlier, as reviewed in the paper by Friedmann, Belletti, Rizzi paper discussed in what follows. The interested reader should go there for references.

[2] Julie Van Dyke and Brian McElree (VD&M) wrote another paper that I found helpful (here). It tries to zero in on a more exact specification of the core properties of content addressable memory systems. The feature that they identify as key is the following:

The defining property of a content addressable retrieval mechanism is that information (cues) in the retrieval context enables direct access to relevant memory representations, without the need to search through extraneous representations (164).

In effect, there is no cost to “search.” Curiously, I believe that VD&M get this slightly wrong. CA specifies that information is called in virtue of substantive properties of its contents. This could be combined with a serial search. However, it is typically combined with a RAM architecture in which all retrieval is in constant time. So general CA theories combine a theory of addressability with RAM architecture, the latter obviating costs to search. That said, I will assume that both features are critical to human memory and that the description they offer above of CA systems correctly describes biological memory.

[3] In other words, unifying RM with CA systems would effectively treat RM effects as akin to what one finds in self-embedding structures. These are well known to be very unacceptable despite their grammaticality (e.g. *that that that they left upset me frightened him concerned her).

[4] My other favorite is the A-over-A principle. Happily, these two can be unified as two sides of the same RM coin if constituents are labeled. See here for one version of this unification.

[5] In fact, you might enjoy comparing the pix in VD&J (p. 197) with (1) above to see how close the conceptions are.

[6]  VD&J do not report whether replacing both D-NPs with pronouns or quantifiers reintroduces the Subject Advantage that replacing barber eliminates. The prediction would seem to be that it should on a straightforward reading of RM. Thus, someone who you/he saw should contras with someone who saw you/him in roughly the same way that (2) and (3) do. VNP (see below) report some data suggesting that quantifiers might pose separate problems. VD&M reporting on the original Gordon & Co studies note “their data indicate that similarity-based interference occurs when the second noun phrase is from the same referential class as the first noun phrase, but it is reduced or eliminated when the noun phrases are from different classes (158).” This suggests that the SBI effects are symmetric.

[7] The scare quotes are here for in the relevant examples exploit a “what makes sense” metric, not a type measure. All the listed expressions in boat-sail and boat-fix examples are of the same semantic type, though only boats are “sailable.” Thus it is really the semantic content that matters here, not some more abstract features. VD&M review other data that points to the conclusion that there are myriad dimensions of “similarity” than can induce SBIs.

[8] VD&J cites the work of Gordon and collaborators. They do not link the abatement of SBIs to syntactic shape but to their semantic functions, their “differeing referential status.” This could be tested. If Gordon is right, then languages with names that come with overt determiners (e.g. Der Hans in German) should, on the assumption that they function semantically the same as names in English do, obviate SBIs when a D-NP is head of the relative. If MR is responsible, then these should function like any other D-NP nominal and show a Subject Advantage.

[9] This is the main point. Of course there are many details and controls to worry about which is why AARO is 50 pages rather than a paragraph.

[10] This might create trouble for the strong RM effects, like moving adjuncts out of WH islands: *How did you wonder whether Bill sang. This is a really bad RM effect. This is correct and the question arises why so bad? Dunno. But then again, we currently do not have a great theory of these ECP effects anyhow. One could of course concoct a series of features that led to the right result, but, unfortunately, one could also find features that would predict the opposite. So, right now, these hard effects are not well understood, so far as I can tell, by anyone.

[11] Note the very tentative nature of this remark. Are there any results from language processing/production/acquisition that implicate purely linear relations? I don’t know (a little help would be nice from you in the know). The NPI stuff and the agreement attraction errors are not entirely insensitive to hierarchical structure. Maybe this: VNP cite work by Friedman and Costa which shows that children have problems with crossing dependencies in coordinate structures (e.g. The grandma₁ drew the girl and t₁ smiled). The “crossing” seems to be linear, not hierarchical. At any rate, it is not clear that the psycho data cannot be reinterpreted in large part in hierarchical terms.

[12] However, from what I can tell, pure linear effects (e.g. due to decay) are pretty hard to find and where they are found seem to be of secondary importance. See VD&J and VD&M for discussion. VD&J sum things up as follows:

…interference is the primary factor contributing to the difficulty of integrating associated constituents…wit a more specialized role arising for decay…

[13] One other concern might be the following: aren’t grammatical restrictions categorical while performance ones are not? Perhaps. Even RM effects lead to graded acceptability, with some violations being much worse than others. Moreover, it is possible that RM effects are SPI effects that have been “grammaticized.” Maybe. So RM is a grammatical design feature of G objects so that such objects mitigate the problems that CA memory necessarily imposes. I have been tempted to this view in the past. But now I am unsure. My problem lies with the notion “grammaticization.” I have no idea what this process is, what objects it operates over, and how it takes gradient effects and makes them categorical. At any rate, this is another avenue to explore.

[14] There are some syntactic implications for the unification of RM and SBI effects in terms of the structure of CA memory. For example, if RM effects are due to CA architecture then issues of minimal search (a staple of current theories) are likely moot. Why? Well, because as observed (see note 2), CA “enables direct access to relevant memory representations, without the need to search through extraneous representations.”

In other words, CA eschews serial search and so the relevance of minimal search is moot if RM effects are just CA effects. All targets are available “at once” with none more accessible than any others. In other words, no target is further/less accessible than any other. Thus if RM special case of CA then it not search that drives it. This does not mean that distance does not matter, just that it does not matter for search. Search turns out to be the wrong notion.

Here’s another possible implication: if decay is a secondary effect, then distance per se should not matter much. What will matter is the amount of intervening “similar” material. This insight is actually coded into most theories of locality: RM is irrelevant if there is only one DP looking for its base position. Interestingly, the same is true of phase based theories, for structures without two DPs are “weak” phases and these do not impose locality restrictions. Thus, the problems arise when there are structures supporting two relevant potential antecedents of a gap, just as a theory of minimality based on SBI/CA would lead one to suppose.

More on experimental syntax

In a recent post (here), I discussed a really nice result that our experimental syntacticians (ES) have delivered wrt island phenomena. As I noted, they have shown that ES can find (and has found) a distinctive super additivity signature of island violations even in highly acceptable sentences in Scandinavian (i.e. independently of overall acceptability). This strongly challenges (IMO debunks) the long held view that islands are parametric, a view that makes little sense when viewed through the lens of the Poverty of Stimulus (PoS) argument. Given that a well-formed PoS argument is virtually apodictic, this is what we should have expected all along. Given ES tools, our expectations have been realized. Chalk this one up as a victory for PoS logic.

I also remarked on another nice feature of this application of ES methods; it has yielded a really novel result. We actually could not “see” island effects using more conventional binary judgment (* vs ok) methods and this misled syntacticians. In other words, opening one’s mind to the conceptually impossible (i.e. “results” that violate PoS logic), encourages (a comfortable, yet fraudulent) skepticism regarding our best-grounded insights (e.g. concerning the universality of island effects) and this in turn leads research in the wrong direction. Well-grounded results, like Ross’s islands should be fixed points in ongoing inquiry and they cannot serve this function if GGers doubt their standing. At any rate, hooray for Kush, Londahl, Sprouse and Almeida.  They have done something both valuable and new.

I wanted to offer up two other results based on ES methods that I believe are “novel” in the sense of being heretofore less easily investigated without ES tools.  I was involved with two of these and so I know something about them. Hence what follows has the whiff of self-promotion. Believe me when I tell you that my contribution to both efforts was very minimal.  I invite readers to inform us of other new kinds of results using ES methods.

The two I know about are both chapters in the book that Jon Sprouse (with a small amount of help from moi) edited (here). The first is the chapter by Johannes Jurka on subject islands in German.  This paper examines these uisng ES methods. It shows that extracting out of subjects is worse than extracting out of objects and (this is the fun part) that extracting out of non-agreeing specifiers lies midway between the two. Thus, extracting out of specifiers is harder than doing so out of complements even in the absence of much agreement or any evidence of displacement.[1] This effect appears to be independent of freezing, which Jurka notes seems to function as an independent factor. Indeed, extraction out of external arguments is always harder/worse than extracting out of complements regardless of movement or agreement. This is very much worth knowing for, if correct, it suggests that subject islands cannot be reduced entirely to freezing effects. In fact, one can go a little further: so far as I know there are not many current theories (as opposed to Barriers style accounts that might be able to accommodate this asymmetry via differential L/theta-marking) that predict that extraction out of a specifier per se should be harder than extraction out of a complement.[2] Thus, Jurka’s ES empirical work suggests that we need to think quite a bit harder about the subject side of CED effects.

The second paper is one that I co-authored with Brian Dillon (again, he did most of the work). The paper presents ES evidence that what islands are sensitive to is syntactic structure.  “Opposed to what?”, you might be wondering. Well, opposed to semantic structure in particular. There once was a time when people tried to reduce island effects to semantic ones (I faintly recall Rodman trying to do this in the early 80s). And given the close connection between semantic hierarchical structure and syntactic hierarchical structure it is very hard to tease apart whether island effects are actually syntactic.  In fact, doing this requires holding the semantics constant and manipulating only the syntactic form, and this is not easy to do while keeping all other factors more or less the same. The above paper does this by focusing on extraction form small clauses and their semantically (largely) identical nominal counterparts. Thus pairs like (1a,b) mean the same thing (both denote events with the same John/Mary participants) but it turns out that extracting out of the nominal complement is harder than extracting out of the small clause complement.

(1)  a. Mary heard John clumsily attempt to kiss Mary

b. Mary heard John’s clumsy attempt to kiss Mary

There are, of course, all sorts of manipulations required on this basic theme to control for all sorts of things (e.g. definiteness effects among others (btw, Brian did this)), but the basic result is that extracting out of nominal event denoting complements is harder than extracting out of their small clause counterparts and this seems to be entirely due to the fact that one is nominal and the other is not.  Again, many of the judgments are in the acceptabl-ish territory so we have a kind of subliminal island effect. At any rate, to my knowledge this is one of the first attempts to pin down the claim that islands are syntactic effects, i.e. effects sensitive to syntactic structure.  In this particular case what matters is the distinction between a nominal and a sentential complement (i.e. labels seem to make a difference here).  If this is correct, then island effects are syntactic phenomena at least in the sense that the relevant primitives need to allude to the syntactic features of constituents.

Again, I mention these papers because they try to do something new with ES methods, they try to find effects that are hard to spot using the easier more convenient ask-your- next-door-neighbor methods. Let me repeat, lest this be grossly misunderstood, that I am NOT endorsing the view that we all now do ES experiments to ground our data. This is not necessary in general (again as Sprouse and colleagues have argued successfully IMO). However, there seem to be times when ES methods can yield new insights, and when this is so, we should not be reluctant to use these (more expensive) methods. When might this be? 

I suspect that it is not an accident that ES has been most successfully applied to island phenomena. Why? Because we know a hell of a lot about islands, both empirically and theoretically. Much of this knowledge is based on data collected in the standard way, and the conventional methods have clearly proven to be very productive.  However, what the ESers have shown is that when we get down to more refined and filigree issues especially in areas that we know a lot about, it should not be surprising that we might need more careful empirical probes.[3] In fact, as I’ve mentioned elsewhere, what we should find surprising is that the very crude methods we have used till now have proven to be so robust and subtle. This surely tells us something about FL, namely that it leaves very deep footprints so that even slapdash methods suffice to probe it. However, we should never have expected this charmed state of affairs to continue forever. Happily, ES, which is pretty easy to deploy, provides another method for probing structure. Happily it largely leads to the same results in the well-understood cases. Happily, it sometimes delivers new insights.  All in all, this is all a very happy fact. So, be happy and be catholic in your choice of tools.

---

[1] An interesting feature of Johannes’ results is that they provide an argument for binary branching. How so? Well, IOs and DOs cannot both be complements given his results. This should be possible were non-binary branching possible. Though I believe that binary branching is in fact a condition on constituency, there are not all that many arguments in its favor, so far as I know. There are the Barss-Lasnik data, bit aside from that I don’t know of many others. Do you?

[2] The one that I do know of (and that Jurka cites) is Uriagereka’s version of multiple spell out, which, to my knowledge is not widely investigated or accepted.

[3] Akira Omaki has used ES methods to explore a Relativized Minimality approach to WH-islands (here). I hope to discuss his results in a forthcoming post on RM. But no reason for you to wait to read it.

Some things to read

I’ve recently run across three little articles that GGers should find interesting (Thx Colin).

The first features David Poeppel. He discusses some recent work by Edward Chang at UCSF who has discovered neurons that respond to “basic “phonemic features,” rather than to phonemes themselves, which are larger chunks of sound.” As David points out in his remarks, this is not news to linguists. But it is a nice example of how what we have discovered, using pretty conventional analytic methods within phono, have mapped pretty directly onto basic neuro units. So when someone asks how linguistics is relevant to the brain sciences, this is a nice case to pull out and wave about. One more thing caught my eye: the relation between brain and cognition is here executed at the level of primitives. As Chang put it, features are the “building blocks of speech and language.” It makes sense to think that the easiest bridges to the brain will be from our primitives to theirs. The reason is that once one gets by this very simple stage complexity grows rapidly and the mapping likely becomes more and more obscure. This recalls the Embick and Poeppel observations concerning the Granularity Mismatch Problem (see here for some discussion). If GG is to make contact with neuroscience (which, IMO, it had better do or risk irrelevance, or worse), it will need to find ways if breaking the complexities of grammar into simpler and simpler sub-units and primitive operations. Happily, GG has been doing this since the mid 80s when constructions were broken down into more primitive operations like ‘moe alpha.’ As all of you know, this is also the main animating conceit of the Minimalist Program. And if what we see above is indicative, this is a very good thing for it’s at the level of simple units and basic operations that we might hope to find links to the neurosciences.

Marcus continues this theme in two pieces. The first is a piece in the NYTs (here) in which Gary argues that we need to return to the idea that the brain is a kind of computer. Apparently, this idea is no longer in fashion. Gary explains why the reasons for dumping this analogy are very weak and actually counterproductive. As he points out the right question is not whether the brain is (like) a computer but what kind of computer the brain is like. This seems absolutely right. The brain is an information processing device. Thus, it computes.

Gary rebuts the standard reasons for ignoring this analogy between brains and computers. But more interesting still he provides an interesting proposal for what kind of computer the brain might be. In a paper with two colleagues (here), he suggests that the brain is very like a “FIELD programmable gate array.” What are these? Well, they

…consist of a large number of “logic block” programs that can be configured, and reconfigured, individually, to do a wide range of tasks. One logic block might do arithmetic, another signal processing, and yet another look things up in a table. The computation of the whole is a function of how the individual parts are configured. Much of the logic can be executed in parallel, much like what happens in a brain.

In the Science piece, Gary & Co propose (p. 552) that neuroscience ought to be looking not for “a single canonical circuit” but for

…a broad array of reusable computational primitives- elementary units- of processing akin to basic sets of instructions in a microprocessor- perhaps wired together in parallel…

This fits with the “basic units/operations” theme I alluded to above. Again, it is at this very basic level that we can rationally hope to make contact with the brain sciences.

One more observation: the Science piece reads a little like a minimalist manifesto for the neurosciences. Most of cognition is made up of several simple operations that link together in different ways for different ends. Thus we should expect to find the same primitives again and again across cognition. This is music to a minimalist’s ears: most knowledge of language resides in recycled basic cognitive circuits. If there is something linguistically special (and I believe that the evidence to date is that there is) then it is a pretty small addition to this basic inventory of primitives. Our job is two fold: (i) to identify that special linguistic operation and (ii) to show that all the rest of G competence can be reduced/analyzed in terms of the remaining cognitively general ones. Conceptually, this program in linguistics fits snugly with the one that Marcus & friends outline for cogneuro, and that is just great!

One really last observation: Gary & Co end their Science piece with the following inspirational peroration (my emphasis NH):

Neuroscience must develop precisely the sorts of experimental tools, detailed brian maps, and computational infrastructures that today’s brain initiatives aim to support, but also a new set of intellectual tools for understanding how, even in principle, systems might bridge from neuronal networks to symbolic cognition…

In other words, it ends in praise of research on how things could work as well as how they do work.  I could not agree more, not only in neuroscience, but in linguistics as well. The obsession with the actual is often a barrier to intellectual progress. For the big questions, it is indispensible. Sadly, IMO, it is often devalued. But you already know that I think this. I find comfort in finding that I am not alone.

Open source journals?

Adam Liter send me this interesting piece on the Dutch university boycott of Elsevier publications. The complaint is one we are all familiar with: too expensive. It seems that academic publishers make lots of money on journals. Moreover, they do this on the back of lots of free labor. Reviewers are free, editor salaries are paltry, and buyers have little bargaining power. Not surprisingly this leads to large returns.  With the internet, the thought that there must be a better way always comes bubbling to mind.  So is there?

The obvious reply is that open source journals should be able to step into this breach and provide the same product at a better price. However, these journals have not taken over. Why not? Well, one reasons is that even such journals need cash and if it is not to be extracts from university libraries then it will come from this who want to publish. Some of this can be offset with grants. But this then means that scholars without such money will have to pony up (and the costs discussed can be large, running into 4 figures).  So, what to do?

I really don't know, but I think that this is an important question, especially for linguists. IMO, we don't have enough good publishing venues and it takes forever for research to get into print. Maybe this is less important than it used to be given things like the Linguist List.  Maybe publishing in LI or NLLT or NLS is not that important intellectually and that the only serious service journals provide is vetting of candidates for tenure. I don't know. What do you think? Is there a dark side to open source journals? Can they replace the large academic publishers? Should they? Your turn.

Pullum on the competence-performance distinction

In an earlier post I discussed a nice little piece by Geoff Pullum (GP) on Aspects after 50 years (here). GP has a second interesting post (here) that discusses a central distinction highlighted in Aspects; the competence-performance distinction (CPD). As he rightly notes, this distinction has caused endless cognitivist’s dissonance. Like GP, I don’t really understand why. It’s a distinction that is rampant, for example, in statistical conceptions of cognition. The trivial distinction between the structure of the hypothesis space and the actual distributions over that space is a version of the CPD. Nobody takes this to be a controversial distinction.[1] If so, we can take a theory of the hypothesis space (what’s its geography?) to be a theory of competence and a theory of how this space gets filled probabilistically to be a theory of performance. This common distinction then provides a first okish pass at the CPD. It is not perfect, but it gets you a lot of the way there.

Another useful explication is that a theory of competence aims to limn the limits of the possible. In syntax we aim to describe the possible sentences of a language in terms of the unbounded products of a given G. We aim to describe the possible Gs in terms of the products of FL/UG. We aim to describe the possible FLs… The sentences/ Gs/FLs we actually encounter are points in a space of sentences/Gs/FLs that could exist. Theories of what could be are theories of competence.

Ontologically, theories of competence ground theories of performance. This is why the CPD is rife with Empiricist/Rationalist (E/R) baggage, which chapter 1 of Aspects elaborates on.  This is also why, IMO, it has proven to be so difficult a concept to get across. Empiricists treat competence as a conceptually derivative notion. First there is performance. Competence, on the E view, is smoothed performance, performance with the variance squeezed completely out, non-noisy performance.  This, however, is not what the CPD demarcates. It is essentially a Rationalist notion, pointing to hidden structure, which actual linguistic items can be used to reveal.

I mention this because, this seems quite different from the way that GP frames matters in his post.[2] For GP, following Edward Lorenz, competence is “what you expect,” and performance is “what you get.” He then elaborates this in terms of what we will find if we squeeze out all the “sporadic and unintended mistakes” of performance. So competence is cleaned up performance. But it isn’t. For example, many many G products will never be performed, (im)perfectly or otherwise, yet they are explanada of a competence theory of G. Humans may fail to internalize many many Gs, yet these Gs may be possible and Gs that at theory of FL/UG should permit. So, what we see, even in statistically cleaned up form (i.e. with all the nosiy variance squeezed out) is not what competence is about.

Of course, we hope that we can get a window into the possible by investigating what is actually deployed. Thus, actual performances are the source of our linguistic data. What people say, how they judge things we present to them, etc. These data are used to plumb the limits of the possible. And some data is perhaps better than others for this purpose. So, if you are interested in what linguistic knowledge consists in (i.e. knowledge of a G) then some data might be better suited than others for probing this. What kind? Well, those that do not run afoul of factors such as limited memory, slips of the tongue, etc. that we think might confuse matters.  So, we clean up the relevant data we use to plumb G competence.  These gussied up linguistic objects are not themselves the targets of explanation. Rather they act as probes into the structure of a G or of FL/UG. In other words, the target of a theory of competence in linguistics (i.e. what we want a theory of competence to explain the properties of) is a G or FL/UG. The intuitions we harvest and the utterances we track are what we use to investigate these structures. The point that Chomsky makes in Aspects concerning the ideal speaker-hearers simply adverts to the fact that some data (e.g. those that abstract away from performance factors like memory and attention limitations) are nosier than others and so less useful as probes of the G and FL/UG systems. In other words, Aspects makes the undeniably correct point so some data are plausibly more useful than others if one’s aim is to discern the structure of linguistic knowledge.

GP interprets the discussion in Aspects quite differently. He understands Chomsky to have been proposing that the “subject matter” of linguistics was “speaker intuitions about sentence structure.” These intuitions are purer than linguistic performances (e.g. utterances) in that they abstract away from (and here GP quotes Chomsky)  “grammatically irrelevant conditions as memory limitations, distractions, shifts of attention and interest, and errors (random or characteristic) in applying his knowledge of this language in actual performance.” However, even if we grant this (which we should), this does not make these intuitions the subject matter of linguistics. No! These intuitions are simply data, evidence that linguists can use effectively to understand the structure of Gs and FL/UG. Chomsky does claim that such intuitions are rich sources of information about the structure of G knowledge and might suggest that they are superior data to corpora data (i.e. recorded speech of actual utterances). Moreover, he clearly thinks that we should dump the prejudice against speakers’ judgments that behaviorism saddled us with. However, Chomsky does not argue in Aspects (so far as I can tell) that intuitions are epistemologically or ontologically privileged data, just that they don’t suffer from some of the problems we might think would mislead investigation.

This is almost certainly correct, and it is not the same as suggesting that these data are the subject matter of linguistics. The subject matter, (aka aim of inquiry) is, as Chomsky puts it, to describe “the mental reality underlying actual behavior.”

I should add that the idea that intuitions are privileged in some way leads to all sorts of misconceptions that psychologists then spend so much time lecturing linguists about. Such judgments are themselves complex performances, with all the problems that performances entail. The most that can be said for intuitions is that have proven to be excellent probes and that they are very stable, robust, and easy to gather. These are important virtues, but they don’t argue for intuitions as such being privileged, as would be the case were intuitions the subject matter of linguistics.

In sum, if I got GP right here (but see note 2), then I think he gets the CPD wrong. What he presents is the Eish version of the CPD. Aspects takes an irreducibly Rish conception of the aim of inquiry and the CPD is intended to highlight this approach. If this is right, then the main problem with getting others to understand the competence/performance distinction lies in them getting to see how closely it is related to an Rish conception of mind (and science actually (see here)). But Rationalism is largely anathema to many practicing neuro-cog types and so the distinction is hard for them to understand (and accept). This is to be expected. On an Eish conception, the most one can make of the notion lies in the quality of data. It marks a distinction between types of data: Performance data is “messy” “competence” data is not. The latter is privileged for that reason. However, this is not the point that the CPD is intended to highlight. It highlights the difference between the products of an underlying mechanism and the mechanism itself. In other words, it highlights the claim that the subject matter of linguistics is Gs and FL. In other words, the CPD carries within itself the project of modern Generative Grammar, and that’s why understanding it is so very important.

---

[1] See the Amy Perfors quote here.

[2] I say “might be” because it is possible to read GP as making the same point as I am making here. If so, great. However, there is another reading where he sees competence as non-noisy performance. But, I may be misreading him here and if so my sincere apologies to GP. That said, the two ways of interpreting the CPD are important so I will continue putting my own construal on GP’s elaboration.