Can Neural Nets do Meaning?

The pandemic has been hard on many of us. It has been a long time since I traveled to an in person conference, or blogged about my experiences. The plan is to create a blog post for each of the three days, but let’s see– I am a little out of practice. Today I concentrate on the invited panel on computational semantics. There were other talks in the main session today but they will have to wait for another blog post.

The day started with a panel discussion on computational semantics. See the listing on the programme here. The three invited speakers, it turned out had different research goals, which was interesting, and I wonder how representative it is of the field. The question I posed to the panel after their (all very) interesting talks, was whether they considered themselves to be pursuing the goal of making the performance of computers on language related tasks better because it would lead to better functionality in various applications, or whether they were interested in modeling meaning tasks computationally in order to understand the human mind better.  Marie Catherine de Marneffe said she was was unequivocally in the former camp, Aaron White in the latter, while Ellie Pavlick was somewhere in transition— she started off being more interested in the former kinds of problems but was getting increasingly interested in the latter.

De Marneffe was interested in getting  computers to perform in a human like way with respect to judgements about speaker commitment to the truth of certain embedded propositions. As is well known, the new deep learning systems, trained on mountains of data (available for languages like English), end up doing stunningly well on standard benchmarks for performance. The speaker commitment judgement is no different, performance is strikingly good. The neural network gets given  simple parametric information about the lexical embedding verb (whether is is factive , or whether it lexical entails speaker commitment in principle), but also gets exposed to the distributional data, since linguistic context such as the presence of negation and other embeddings are necessary to make the judgements in question.  It turns out that these kinds of neural networks perform extremely well for example on neg raising contexts, generating human equivalent judgements for sentences like

I don’t think he is coming.

 However, there are a few kinds of sentence where the neural networks fail spectacularly. These are instructive. Two examples from the talk are given below, with the clause for which speaker commitment judgement fails shown as underlined.

(1) I have made many staff plans in my life and I do not believe I am being boastful if I say that very few of them needed amendment.

(2) I was convinced that they would fetch up at the house, but it appears that I was mistaken.

De Marneffe pointed out these examples and speculated that the problem for the neural nets is pragmatics and/or real world knowledge. (2) is striking because even the smallest most ignorant child would get this one right, so it seems to show  that whatever the neural net is doing, it really is not doing anything remotely human like. Maybe having a real embodied life and connections to  truth in the world is necessary to fix (2). But the problem with  (1)  seems to me not to be not so much about pragmatics as about embedding and hierarchical structure, which the neural net simply is not tracking or using as part of its calculation. Personally, I think the `problem’ with pragmatics, in terms of inferential strategies is overstated. I am pretty sure you can teach neural nets some inferential algorithms, but compositional structure and real grounding for meaning both seem to be the real sticking points.  But we only see this in cases when the linear distance cooccurrence data is non informative of the actual meaning.  It is sobering to notice how seldom those cases actually come up, and how often the simplistic heuristic delivers as a proxy for the more complex reality. How worried you are about the existence of these examples really depends on which of the two issues outlined above you are trying to solve.

With regard to Being in the World, Ellie Pavlick presented her work on trying to teach meaning grounding to neural nets, as a way of probing whether such training on physical properties of events denoted by motion verbs would help in acquiring the right behaviours and underlying representations. The evidence seems to be that modest gains in performance are indeed possible in certain domains based on this kind of training.  But here one wonders whether we can follow up those gains in all other domains without fully recreating the learning environment of the child in all its gory and glorious detail. The reductio of this approach would be a situation where you require so much data and nuance that it would be impossible to construct short of birthing your own small human and nurturing it in the world for five years.  As Ellie rightly pointed out in discussion however, the great advantage and excitement of being able to program and manipulate these neural nets is the controlled experiments you can do on the information you feed it, and how you can potentially selectively interrogate the representations of a successful model to try to come up with a decomposition of a complex effect, which might in the end be relevant to understanding the cognitive decomposition of the effect in humans.

Aaron White’s talk was on an experiment in training a neural net to match acceptability ratings leading to the induction of a  type structure for different constructions. The basic model was a combinatory categorial grammar with standard basic types and modes of combination. The intermediate interchange format was vector space representations, which are flexible and don’t require prejudging the syntax or the compositional semantics. The point of the training is to somehow see what gets induced when you try to create a system that best predicts the behavioural data. The test case presented was clausal embedding,  and peering under the hood afterwards, we can ask what kinds of `types’ were assigned to clausal complements of different varieties, and with different embedding verbs.  The types induced for clausal complements were very varied and not always comprehensible. Some seemed to make sense If you were thinking in Inquisitive Semantics terms, but others were harder to motivate. All in all, it seems like the job of interpreting why the model came up with what it did is as hard as the original problem,  and moreover bearing an ill understood and equally complicated relationship to the original problem of how humans `do’ meaning composition.  There are a lot of details that I clearly do not understand here.

All in all, it was a fascinating panel raising a lot of big picture issues in my own mind. But I come away with the suspicion that while BERT and his descendents are getting better and better at performing, their success is like the equivalent of getting the answer 42 to meaning of Life the Universe and Everything. It still does not help if we don’t know what exactly their version of the question was.  

Here is a short blog version (without slides) of a talk I gave at the recent Jabberwocky workshop hosted jointly by UMass Amherst and the University of Bucharest (thank you Camelia Bleotu and Deborah Foucault for a great initiative!). The ideas in this talk were rather non-standard and I suspect rather unpopular, but the concept was interesting and it was a great group of people to potentially interact with. Unfortunately, the time zone and weekend timing of the workshop did not allow me to participate as fully as I would have liked, I am airing those ideas here on this blog just in case someone is interested.

Jabberwocky sentences consist of syntactically well formed sentences with nonsense content words like this one I just made up: She didn’t glorph their lividar

If you are a syntactician, the nonce words here are a clever way to eliminate the effect of real lexical items and conceptual content, and zero in on combinatorial processes which underlie sentential structure and generativity. The very fact that we can make these sentences, seems to show that this aspect of language is distinct and modularizable away from the Lexicon per se. It is good to be able to abstract away from contentful LIs in a variety of methodologies, because controlling for frequency, semantic prediction, association etc. can be hard. From the point of view of syntactician, Jabberwocky sentences seem to offer a way of surgically removing the messy bits and to target pure syntax.

So the lexicon is hard, but in modern Chomskian views of grammar, the Lexicon is also the boring bit, where memorized chunks exist, but where no generative processes reside. This is taken to extremes in the Distributive Morphology tradition, where roots are devoid even of syntactic information that would tell you how to insert them in a sentence. The formal semanticists tend to concur: in that tradition we prove theorems of the form: Snow is white is TRUE iff ‘snow is white’ (Davidson 1967). Where the contentful lexical items are simply repeated in the metalanguage, languishing there for someone else (presumably NOT the formal semanticist) to elucidate.

However, there are some reasons to be a little suspicious of the possibility of excising the LI in a clean modular fashion.

Jabberwocky and fMRI

Fedorenko et al. (2010) develop a localizer task for helping in the analysis of regions of interest (ROIs) for linguistic experiments using fMRI. They use four conditions:

1. Sentences (The Sentences condition):

2. Scrambled Sentences (Word list condition):

3. Jabberwocky Sentences:

4. Scrambled Jabberwocky Sentences (the Non-words Condition):

Sentences > NonWords showed the language regions. Words and Jabberwocky both showed intermediate activation of the sentence regions but could not be reliably distinguished from each other. Words > NonWords and Jabberwocky > Nonwords showed ‘inconsistent and variable results across subjects’. This is disappointing if we think that jabberwocky sentences should show the brain doing its pure syntactic thing.

Jabberwocky Sentences and Neural Oscillations

There has been recent work in neurolinguistics exploring the idea that the processing of hierarchical linguistic structure is correlated with the synchronization of brain rhythms in various frequency bands. Kaufeld et al. (2019) (2020) recorded (EEG) while 29 adult native speakers (22 women, 7 men) listened to naturally spoken Dutch sentences, jabberwocky controls with morphemes and sentential prosody, word lists with lexical content but no phrase structure, and backward acoustically matched controls.

I quote: “Mutual information (MI) analysis revealed sensitivity to linguistic content: MI was highest for sentences at the phrasal (0.8–1.1 Hz) and lexical (1.9–2.8 Hz) timescales, suggesting that the delta-band is modulated by lexically driven combinatorial processing beyond prosody, and that linguistic content (i.e., structure and meaning) organizes neural oscillations beyond the timescale and rhythmicity of the stimulus.´´

The jabberwocky sentences on the other hand were no different from word lists with lexical content and no phrase structure on this measure.

One reaction to this kind of disappointing result is to say that syntax is just not really modularizable the way we thought. This seems to be the position of Blank and Fedorenko (2020), Mahowaldi et al 2022, essentially embracing work in Construction Grammar (Goldberg 1995, Goldberg and Jackendoff 2004).

These kinds of authors are also quick to point out that we don’t ‘need syntax’ to understand complex sentences most of the time, since lexical content and real world knowledge do the job for us. These ‘fake debates’ present us I think with a false set of analytic options. Grammar is not all constructions with lots of rich lexical content interacting with statistical properties of the world, nor is it Super Syntax the heavy lifter (oh wow recursion) with lexical items a relic of fuzziness that syntax can mold and structure for its creative purposes.

My own interpretation is to say that syntax exists (and is a cool weird thing), but that it needs to feed off content for the whole engine to get rolling. this means that our job as linguists requires us to understand (at least) two things:

(1) What are lexical meanings?

(2) How do they integrate with syntax in a compositional and combinatorical way?

So, we should use Jabberwocky sentences not to erase the lexical item, but as a way of trying to understand it better. All words are nonce before we know them.

The Real Beast: Jabbo Sapiens

This talk is a plea to use Jabberwocky sentences and nonce words to help usunderstand not the comprehensible residue, but the things they are replacing—- content words themselves! These little monsters, these Jabbo Sapiens turn out to pose loads of hard problems for compositional semantics and understanding how we communicate with other minds.

One might argue with Donald Davidson that building truth theorems is already hard, and good enough, and that it really is not the immediate job of the formal semanticist to elucidate the meanings of the individual lexical concepts snow and white.

The problem with the meanings of open class lexical items:

(i) they are conceptually polysemous while still being atomic with respect to how they function within the system , and

(ii) they undergo productive compositional processes with each other.

The latter point shows that understanding their behaviour is an important component of understanding the central properties of the human language system and its powers of productive meaning generation.

The psycholinguistics literature is very clear at showing us that there is a hub, or unity to the lemma with a localized point of access. This point of lexical access seems to be in the mid temporal gyrus (MTG) is independent of whether the sensory input is visual or auditory (Indefrey and Levelt 2004, Hickok and Poeppel 2007. Friederici 2012). Activation in this area can also be tracked using MEG and fMRI. Based on both neurolinguistic and behavioural evidence, we have strong support for the existence of the lemma which is the lexeme family underlying a symbol and all of its inflectional forms. Specifically, we know that lemma frequency as a whole (not the frequency of individual forms) modulates effects in the 300/450 ms time window in the MTG (Solomyak and Marantz 2010).

This literature is important because it shows that there is a lemma hub for all inflectional forms of the ‘same’ lexeme. But what constitutes ‘sameness’ in this sense? While in practice, it is not always easy to decide whether a pair of meanings associated with a form are homonyms or polysemic variants, or what leads learners/speakers to classify them as such, the evidence now seems clear that we can distinguish between cases where there must be two ‘lexical entries’ versus cases where there must be one. The cases where we have clear evidence for one lexical entry involve lemmas which characteristically embrace a large number of polysemic variants. Thus, polysemy is sharply distinguished in terms of cognitive consequences from homonyny, or genuine ambiguity, in which two distinct lemmas happen to share the same form. Polysemous readings are bunched together for the purposes of priming. Polysemous meanings are facilitory in word recognition, while genuine homonyms are inhibitory and cause slow downs in processing because of more alternatives remaining active. (Rodd et al. 2002 (lexical decision), Beretta et al. 2005 (MEG)),

Jabba Sapiens and Polysemy in Acquisition

How does a learner decide to group forms heard under the same umbrella lemma, the ‘same lexical entry’ if you will. Both the typological evidence and evidence from developing semantic competence in children show that polysemy is natural and ubiquitous. Novel word learning in children shows generalization across polysemous senses, even when the denotational percepts are quite different (Snedeker and Srinivasan 2014). Children also distinguish clearly between homonymy and polysemy at an early age, before they pass any tests of metalinguistic competence, showing that the difference cannot be metalinguistic, as claimed by Fodor and Lepore (2002) (Srinivasan and Snedeker 2011). Moreover, certain types of polysemy seem to be not idiosyncratically memorized, but are plausibly part of a pervasive conceptual system underlying all languages. For example, the container/containee polysemy was found by Mahesh Srinivasan and Rabagliati (2019) across 14 different languages (see also Zhu and Malt 2014 for crosslinguistic evidence).

.The take home point of this blog post is the following. Current formal semantic and morphosyntactic models fall short on explaining how symbolic primes of the open class system are compositionally integrated into sentences. Lexical Items are usually relegated to a sealed off no-man’s land that is somebody else’s business. But how the two domains interact in practice is never made explicit and turns out to be both HARD and IMPORTANT

References

Asher, N. (2011). Lexical Meaning in Context: A Web of Words. Cambridge: Cambridge University Press.

Beretta, A., R. Fiorentino, and D. Poeppel (2005). The effects of homonymy and polysemy on lexical access: an MEG study. Cognitive Brain Research 24, 57–65.

Blank, I. and E. Fedorenko (2020). No evidence for differences among language regions in their temporal receptive windows. NeuroImage 219, 116925.

Fedorenko, E., P.-J. Hsieh, A. N.-C. n on, S. Whitfield-Gabrieli, and N. Kanwisher (2010). New method for fmri investigations of language: Defining ROIs functionally in individual subjects. Journal of Neurophysiology, 1177– 1194.

Fodor, J. and E. Lepore (2002). The emptiness of the lexicon: Reflections on Pustejovsky. In The Compositionality Papers, pp. 89–119. Oxford University Press.

Friederici, A. (2012). The cortical language circuit: from auditory perception to sentence comprehension. Trends in Cognitive Sciences 16 (5), 262–268.

Goldberg, A. (1995). Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press.

Goldberg, A. and R. Jackendoff (2004). The English resultative as a family of constructions. Language 80, 532–568.

Hickok, G. and D. Poeppel (2007). The cortical organization of speech processing. Nature Reviews Neuroscience 8(5), 393–402.

Indefrey, P. and W. J. Levelt (2004). The spatial and temporal signatures of word production components. Cognition 92(1-2), 101–144.

Kaufeld, G., H. R. Bosker, P. M. Alday, A. S. Meyer, and A. E. Martin (2020). Structure and meaning “entrain” neural oscillations: a timescale-specific hierarchy. Journal of Neuroscience  40 (49) 9467-9475.

Leminen, A., E. Smolka, J. D. nabeitia, and C. Pliatsikas (2018). Morphological processing in the brain: the good (inflection), the bad (derivation) and the ugly (compounding). Cortex 116, 4–44.

Mahesh Srinivasan, C. B. and H. Rabagliati (2019). Children use polysemy to structure new word meanings. Journal of Experimental Psychology: General 148(5), 926–942.

Marslen-Wilson, W. T., M. Ford, L. Older, and X. Zhou (1996). The combinatorial lexicon: Priming derivational affixes. Proceedings of the Eighteenth Annual Conference of the Cognitive Science Society 18, 223–227.

Marslen-Wilson, W. T. and Tyler (2007). Morphology, language and the brain: the decompositional substrate for language comprehension. Transactions of the Royal Society of London. Biological Sciences 1481 (362), 823–836.

Pustejovsky, J. (1995). The Generative Lexicon. Cambridge, Ma.: MIT Press.

Rodd, J., G. Gaskell, and W. Marslen-Wilson (2002). Making sense of semantic ambiguity: semantic competition in lexical access. Journal of Memory and Language 46, 245–266.

Sahin, N. T., S. Pinker, S. Cash, D. Schomer, and E. Halgren (2009). Sequential processing of lexical, grammatical, and phonological information within Broca´s area. Science 5951 (326), 445–449.

Snedeker, J. and M. Srinivasan (2014). Polysemy and the taxonomic constraint: children’s representation of words that label multiple kinds. Language Learning and Development 10(2), 97–128.

Solomyak, O. and A. Marantz (2010). Evidence for early morphological decomposition in visual word recognition. Journal of Cognitive Neuroscience 22(9), 2042–2057.

Srinivasan, M. and J. Snedeker (2011). Judging a book by its cover and its contents: The representation of polysemous and homophonous meanings in four-year-old children. Cognitive Psychology 62 (4), 245 – 272.

Whiting, C., Y. Shtyrov, and W. Marslen-Wilson (2014). Real-time functional architecture of visual word recognition. Journal of Cognitive Neuroscience2 (27), 246–265.

Zhu, H. and B. Malt (2014). Cross-linguistic evidence for cognitive foundations of polysemy. Cognitive Science 36.

The term Dark Ages is associated with a period of time in medieval Europe sandwiched roughly between the fall of the Roman Empire and before the Renaissance, and in which allegedly ideas and growth were stultified by fear and superstition, and religious dogma. I am reliably informed that those  early  historians who invented the term (Petrarch) overstated their case and exaggerated this narrative for effect and that in fact lots of good things happened in e.g. 9^th century England, but of all of that I am not personally in a position to judge. And in any case it is not the topic of the present blog post.  Rather, I use the term Dark Ages to refer to the period that we currently find ourselves in—the 21^st century, but more properly starting from the period roughly at the end of the 1990s—  with respect to the realm of Academia.  In fact, the term Dark Ages is actually a gesture towards positivity,  because it implies that it will followed by a period of enlightenment. I certainly hope that is the case, but if so, it does not seem to me that it is imminent.  It is also not lost on me that the phrase Dark Ages conjures up impressions that are rather different  (plague/pandemic notwithstanding) from the shiny, clean, big data, money driven, professionalized academic spaces that our universities are curating.   But bear with me. It’s a metaphor.

The impetus to put some thoughts down on paper came after I received a bulk email from my head of department asking for volunteers from the faculty to join working groups that would collaborate on writing up the first draft of our university’s long term strategy under its new leadership team that had just taken over. I must confess I was curious about the leadership direction my university was going to go in under the new vice chancellor. I felt that under the previous leadership team, our university had moved in all kinds of bad directions ideologically: seeking to run the university more like a business,  taking more and more decisions in a top-down fashion in a more opaque and less accountable way, blindly implementing reports and checks and increased paperwork to measure and assess every quantifiable aspect of success and failure. Maybe the new team would take a fresh approach! So I glanced at the strategy plan that was going to be fleshed out into this important document.

Brave New World?

From the government we inherit a remit to deliver on education and lifelong learning, research, public dissemination, relevance to society and work life, innovation and creation of value. My university in particular promises to pair this with an emphasis on the North and the Arctic (obviously), has a commitment to open access science (you can’t hate it), and promises to use its multiplicity and heterogeneity of campuses, people and skills as an engine for problem solving and creating innovative solutions to the needs of society now and in the future (blah blah blah).    If you trawl around the internet for university strategy documents you will find every university using the very same buzz words, so its hard to figure out what it all means in practical terms, especially the last one. It is also hard to actively disagree with the positive statements in such strategy documents— they all sound like such good things, don’t they? 

In interrogating myself about why I felt underwhelmed and discontented with the outline I read, I decided to force myself to articulate what I thought should be the central remit of a university, and what I thought was missing.  Everyone is emphasizing  relevance, innovation, vision, and  how grounded their university is in their own community’s particular needs and sources of expertise.  It seems to be the zeitgeist. Maybe that is just the sign of  good ideas taking off.  And the language is so vigorous, forward thinking and optimistically engaged (albeit on the vague side).  Maybe I am just being churlish to be picking holes in the vision (`Churlish’ is just one of the kinder words that could be applied to me these days. I embrace it.).  But as I continued to think about it, I came up with at least two major areas where I believe universities have a solemn duty to contribute, but which have completely disappeared from public discourse.

The bottom line is that Universities are being conceived of  on the education side as engines for creating workforces, and on the research side as crucibles for technological innovation. The University is seen as the tool of Capital, and it is funded to the extent that it fuels Growth, and precisely the workforce that the holders of Capital need.   Those making the strategy decisions at the top will tell you that of course this is what the students want too, they want  jobs. Sure. That seems to be the bare minimum though. There are other things that the university in a mature democracy should be doing beyond that bare minimum. But these things have disappeared from strategy documents, or even the strategic thinking of educators and they are actively being eroded because of it.   I give them an airing here.

Education: Critical Thinking and the Challenge of True Democracy

This is clearly relevant for society these days! Just not for getting you a job.

I think democracy is hard and it’s fragile. But it’s the only system worth having and democracies need to work actively to maintain its health. Democracy and voting rely on a nation’s communities having access to education and information, and a sense of responsibility about what is at stake. In most university programmes before the Dark Ages, any degree that was taught by good lecturers provided transferrable skills of critical thinking and assessment of argumentation, and knowledge of how to go about reliably finding out whether something is true.  In the modern age, we have our own special problems concerning how information is disseminated and checked, and in the rise of propaganda and the difficulty of escaping bubbles. Our universities need to take the lead on giving people the skills to be able to navigate the increasingly tricky situation of finding reliable information sources, and also in recognizing bias, learning how to overcome emotion or prejudice in assessing arguments, etc. Our universities also need to take the lead in actively supporting the humanities (by which I mean not just not taking money away but actually channeling money into it). Because every young person needs to understand the cultural and historical context of the world they will be living (and voting!) in, and they need to be exposed to other minds and voices through fiction, which develops empathy and helps transcend tribalism.   (At MIT, where I did my undergraduate education, we had a humanities requirement. Everyone had to take a couple of humanities courses, of their choice, regardless of whether they were majoring in Engineering, or Math or Chemistry or whatever. I think this should be built in to all university degrees).

I am not so naïve to think I will ever convince a country or a university to make this their major remit. But I sure would like to see it as a bullet point in a strategy document.

Education (Lifelong): Tools  for Understanding,  Personal Development and Satisfaction

If you get down to first principles, growth and the economy are not really the things we need as a society. We want everyone to have the necessities of life and to have the opportunities to live full, happy and fulfilled lives. The modern world is  one where we humans have an increasing proportion of time that we can devote to leisure, because hard time consuming physical labour has been taken over by machines and tedious labour by computers.

How do we make ourselves happy?  

Education gives people tools and resources to keep learning, and understanding their world.  This leads to happiness.  Curiosity-based learning, acquisition of new skills,  leads to the appreciation of complex and satisfying forms of leisure and helps us be less bored and passive in our consumption of entertainment.  

Education is not job training. Education is something that human minds thrive on and we do it not to promote growth or technological advance, but just because human minds love to be so engaged.

I would love to rethink the remit of modern universities based around feeding curiosity and developing young people’s skills and resources with the goal of helping them find the thing that they are good at. I am guessing that then, whatever they end up specializing in, they will find a niche in society where they can do a job that contributes to the society’s goals and that makes them happy.   

Research: Curiosity Based Research

When it comes to research, the primary emphasis should be on curiosity driven research. And there should be no competition for grants and funding.

Right now, senior academics as well as early career researchers are forced into an endless cycle of applying for grants and producing publication points. The grants nowadays are skewed towards societal `relevance´ and impact, regardless of whether this fits in with the researcher´s own set of scientific questions. At first, a decade or so ago,  we just added an Impact paragraph to our applications, but now, increasingly, the whole research agenda must be re-thought and new strands of inquiry invented just to get on the grant bandwagon. Primary basic research is not really respected unless it brings in big grant money, and big grant money increasingly depends on subjective decisions of relevance and coolness. The standards of grant applications get higher and higher. Most of these grants are worthy and interesting, and if history of science tells us anything, it is generally impossible to predict what new thoughts or ideas are going to lead to big advances in some body of knowledge. In the mean time, the lottery for what gets funded is driven by forces that are random at best and skewed in an overly superficial direction at worst. And most importantly, researchers´ time is eaten up in this fruitless and soul destroying activity.

I read somewhere that if you take all the money that is spent in organizing the application system, the reviewing system and the administration and reporting of grants across the academic world, you could just give EVERYONE a research grant and resources to pursue a question and save big bucks.

Well, I guess that´s not going to happen. But the system is rotten and we all know it.  It is built on competition and insecurity. Young and early career researchers are experiencing financial insecurity, stress and burnout from increasingly unrealistic expectations,  with very little in terms of intellectual reward.

Climate Change and Man’s Relationship to the Planet

Given the scientific consensus on this, I am sort of surprised that a University like the one I am in,  that wants to be a leader on the North and the Arctic,  does not explicitly come out and say that it wants to lead on helping to reverse the damage we have done to the planet and mitigate the effects of the ongoing climate crisis, especially as some of the clear first signs of polar melt etc. are Arctic issues. Maybe this will come in the details of the strategy document that I am not going to sign up to help write.  But I am not holding my breath. It would probably be considered too political to state such a remit. Although the scientists do seem to agree that these are basic facts, not opinions.  I suspect that the present emphasis on local rootedness is directly connected to universities´ non engagement with issues of a global, universal nature. The world has become increasingly globalized. As long as universities shy away from the big hard questions and see their remit as providing growth, jobs and research grant money to their own local patch, they will not be the engine for critical pushback and change that we so desperately need.

But Some Things Have Got Better, Right?

Since the nineties, some things have improved in certain parts of the world. Rights for LGBT+ trans folks have improved, and diversity in the higher echelons of power in terms of representation of women, people of colour, etc. has improved somewhat (I feel personally that the status of women in academia has stagnated somewhat since the nineties and has lagged behind other kinds of progress).   Access to higher education has improved in many parts of the world.  In many arenas, new, fresh and progressive voices are being heard for the first time above the drone of the wisdom of the perennially entitled.  This is as it should be. As society changes and the people who were not privileged from birth come to have access to education, so will there be changed discourse and re-evaluations and upheaval.  This is also what universities are for. But  I fear that these forces are being managed and de-toothed as we speak. And even access to privilege through education is being clawed back on two fronts— both by turning universities into job making factories, and also by containing and demoralizing its employees who strive to teach and think while getting grants, being relevant,  and preventing their academic areas from getting the axe (is my course popular enough? Does this degree add value in terms of increasing the projected salary of those who take it?). 

The `Dark Ages´ in my long saga refers to our modern era with its commodification of intellectual capital and the Control of Academe by those who currently control the economy.

This piece will no doubt read to most of you who make it this far, as  quixotic, irresponsibly naïve and deeply impractical. But I would remind you that as a GEN X-er (who are they, again?), I actually do remember a time when these things were explicitly talked about in educational circles. So these ideas had not yet vanished from the discourse when I was an undergraduate. And they are not inherently impractical either.  I have watched the narrative shift, continuously and inexorably (just as the political narrative has shifted), to the extent that we have all been made to swallow as a basic premise the idea that anything other than the worklife relevance zeitgeist is untenable (just like the false belief that anything other than the free market and global capital is an untenable system— the two narratives are btw not unrelated).   

Petrarch talked about the Dark Ages primarily in relation to the light that had come before in the form of classical antiquity, not in relationship to the Enlightenment to come. I don’t know if there is any backlash on the horizon that could lead us out of  the stranglehold that this package of ideas has on the world at the moment.  I do not know what it would take to turn this particular boat around. I fear that things will have to get a lot lot worse before they will be allowed to get better.  But I am in the market for ideas!

So far we have argued that the formal semanticists use of an intermediate logical language (the semantic representation) as discussed in earlier posts, is widely considered by the field to be at the level of a computational theory in the sense of Marr 1982, and is not intended to encode internal mental representations of meaning in any psychologically real fashion.

So what understanding of the human mind then do we gain from a study of the syntax-semantics interface construed in this way? The whole enterprise is made more difficult by the fact that we are essentially attempting to solve an equation in three unknowns: we don’t yet know what syntactic representations (their primes and primitive operations) actually look like, we don’t know what those abstract language specific semantic representations look like, and we do not understand the principles of the mapping between the two, except that we know that they must be systematic.

The history of generative grammar shows that there are a multiplicity of different formal proposals concerning what syntactic representations actually look like, with no emerging consensus currently in sight. And we can see from the history of formal semantics as well that the mapping rules change drastically depending on the type of syntactic theory it is interfacing with (cf. Lechner 2015; Partee 2014). The Semantic representation language was taken over from formal logic systems and it too has adapted slowly over time to form a better fit for the syntax (the particular kinds of syntax) that the formal semanticists are mapping from. As the history of syntactic theorizing has shown, there is always a choice between enriching the syntactic representation, or enriching the mapping rules between it and the semantic representation language. Within generative grammars alone, at least two different trends can be distinguished: more derivational and/or abstract syntactic theories whose abstractions in the form of covert rules and implicit structures (the Logical Forms of classic GB syntax, but also abstractness in the form of empty categories and implicit or unpronounced structure) are motivated by generalizations over interpretations; less abstract, more direct and monostratal syntactic representations (e.g. Categorial Grammars, Lexical Functional Grammar, Montague Grammar itself, and Head-driven Phrase Structure Grammar) form the input to mapping rules which in turn must be more flexible and rich in their output possibilities. It easy to see that in this kind of situation, the nature of the mapping rules and the intermediate representations built can be quite different from each other. The primes of the semantic representation language are also subject to variability from the pragmatics side. Every time a proposal is made about a pragmatic principle that can deliver the correct truth conditional results from a more indeterminate semantic representation, this also forces the primes of the semantic representation to be adjusted (e.g. see the effect that Discourse Representation Theory had on the interpretation of the definite and indefinite articles in English). Every time a change occurs in one of these three areas, the details of the whole package shift slightly. The only thing that remains constant is the anchoring in truth conditions. We have to get there in the end, but if these are purely computational or instrumental theories, then we should not put too much stock in exactly how we get there, implementationally speaking. Even compositionality, as a Fregean principle constraining the relationship between syntactic and semantic representations (see Heim and Kratzer 1998) can always be saved by employing the lambda calculus (Church 1936)— a mathematical innovation which allows the decomposition of the logical representation of complex expressions into pieces (higher order functions) that can match the constituents that syntax provides (whatever those turn out to be). So compositionality in this technical sense turns out not to be the criterion according to which these theories can be distinguished from each other. Only if we believe that these convenient semantic representations that we posit have some kind of cognitive or algorithmic reality, or that at least there is some cognitive reality to the boundaries being drawn between the different components, is the specific research area the syntax-semantics interface distinguishable from formal semantics simpliciter. In fact, most formal semanticists are unwilling to do any `neck baring’ let alone `sticking out’, in the area of psychological or cognitive prediction.

Unlike the formal semanticists proper, those of us working at the interface are interested in intermediate representations that we believe bear some organic relation to actual representations in actual minds. For many of us, the quest is to understand how the syntactic system of language discretizes and categorizes in order to create a workable symbolic tool for the mind, while still liaising with the brain’s general, language independent cognitive machinery.

Of special note here is a recent minority movement within formal semantics/philosophy of language towards exploring natural language ontology (Moltmann 2017, 2020). Moltmann in particular has argued that natural language ontology is an important domain within descriptive metaphysics (using the term from Strawson 1959), which is distinct from the kind of foundational metaphysics that the philosophical tradition tends to engage itself in with its spare ontological commitments of Truth and Reference. I see natural language ontology as primarily interrogating our assumptions about the nature of intermediate Semantic Representation that mediates between syntax and truth evaluable representations part, building its primes based on the ontological commitments implicit in natural language(s) itself. As Fine (2017) argues, there is a case to be made that progress in foundational metaphysics relies on a close and nuanced understanding of the descriptive metaphysics involved in natural language ontologies. But even if that were not the case, it seems to me that the project of natural language ontology is crucial if we are to understand the compositional products of meaning and meaning building in language and the mechanisms by which it is embedded in our cognition and cognitive processing more generally. The spare and elegant axiomatization of semantic descriptions anchored just in truth and reference to particulars simply does not do justice to content and partial and incremental contents that we see in language. Exploring natural language ontology in its own right, taking the internal evidence as primary is a prerequisite to getting this kind of deeper understanding. Thus, even though we might think of the syn-sem^E as a computational theory, we can still have the goal of developing a language of primitives on the Semantic Representation side that is more responsive to the implicit categorization found in natural language. Formal semantics took its initial language for the semantic representation from formal logics, but has also repurposed that representation over time to fit natural language better. The research area of natural language ontology takes that goal to its natural conclusion and questions the basic ontology of these representations, and potentially moves the model closer to one that will eventually be more commensurate with cognitive and neurolinguistic theories.

In turn, the patterns that emerge robustly from this kind of natural language investigation, provide clues to both the nature of language itself but to the realities of the cognitive systems that it is embedded in. In part I, I laid out three types of question for I-semantics: Type A questions concerning descriptive generalizations relating semantic systems and the cognitive system they feed; Type B questions related to acquisition and cognitive development; Type C questions concerning the feed back effects of having a language on the very cognitive systems that it subserves. I close this post with a number of examples of phenomena that I think count as instance of Type A generalizations. Note that the existence of these `universals’ would be a surprising fact if general cognition were just one symmetric side of a listed form-meaning pairing. While there seem for example to be no deep generalizations concerning how syntactic primitives are mapped to externalized signals, there are candidates for universals in the mapping to I-semantics. I give some possible candidates in the following list:
(i) Without exception crosslinguistically tense information is represented hierarchically in the syntax outside of causation in the verbal domain, and referential facts such as novelty or familiarity of reference are represented outside of size, colour and substance in the nominal domain (see Julien 2002)
(ii) All human languages make category distinctions within their lexical inventory, minimally N(oun) vs. V(erb) (Baker 2003), and we know that these kinds of syntactic category distinctions cannot be predicted from external facts about the world. But what is this a discretization of in our I-semantics of the world?
(iii) All human languages show open-ended combinatorical ability of open class items to build creative new meanings.
(iv)Semantic modes of combination can be classified minimally into selectional, modificational and quantificational relationships. In other words, even though there is no single semantic combinatoric nexus that will cover all the attested forms of semantic combination, there seems to be a restricted set of semantic nexus types that all languages seem to use (see Higginbotham 1985; Jackendoff 2002) conditioned in systematic ways by syntax.
(v) Quantificational relationships in the semantics always correspond to a particular hierarchical format in the syntax, with the restrictor of the quantifier in combination with the operator, and the scope of the quantifier combined with that. This correlates with the semantic conservativity of all natural language quantifiers (Barwise and Cooper 1981, Lewis 1975).
(vi) The semantics of scalar structure is tracked by linguistic formatives across the syntactic categories of N(oun), V(erb), A(djective) and P(reposition), in all the languages that have been studied.

These are basic empirical generalizations at a fairly abstract level about how human languages compile meanings, and independent of the existence of the Minimalist programme, these are things that it seems to me are the job of the theoretical linguist to pursue in some way. Thus, properly understood, the Minimalist Programme does carve out an interesting and important domain of inquiry, one that might legitimately be called the syntax-semantics interface (Syn-Sem^I).

There are two ways of thinking about the syntax-semantics interface. One involves grounding meaning of sentences in descriptions of the world that make them true and investigating the feeding relationship between syntactic representations and truthmakers in this sense. This is a descriptive enterprise, and gives rise to computational proposals involving language specific abstract semantic representations mediating between the syntactic representation and truth conditions. We could call this Syn-Sem^E. This is not inconsistent with the strong minimalist thesis construed in the way I have argued it can be in Part II, but it is not directly pursuing it.

It is important to reiterate that the formal semanticists use of an intermediate logical language (the semantic representation) is widely considered by the field to be at the level of a computational theory in the sense of Marr (1982), and is not intended to encode internal mental representations of meaning in any psychologically real fashion.

I suspect however, that many people working at what they think of as the syntax-semantics interface think they are doing something different from straight up formal semantics. Unlike the classical formal semanticist, some of these do not use the semantic representational node as merely an instrumental device to mediate between the syntax and truth conditions (seeking accuracy, efficiency and elegance in doing so but no more). They are often also interested in exploring the relation between two systems of representation both of which are internal, and in understanding how the syntactic system of language discretizes and categorizes in order to create a workable symbolic tool for the mind, while still liaising with the brain’s general, language independent cognitive machinery. (I think, for example, that this is what Jackendoff in his work means by Semantics.)

An internalized semantic representation system that is tightly coupled to linguistic representations could be called I-semantics, since it does not represent facts about the world, but is an aspect of the mature I-language system of an individual speaker/hearer (in Chomsky’s 1986 sense in Knowledge of Language).

Thus, the other way of looking at the syntax-semantics interface is in terms of. seeking explanations for the way syntax turns out to be. For this particular set of research questions we are interested in humans’ mental representations of meaning as the mutually determining factor interfacing with syntax. We could call this Syn-Sem^I . The pursuit of the questions concerning this latter interface could be considered a direct pursuit of a strong minimalist agenda. It will inevitably feed off the results of Syn-Sem^E, as described above, but also of the results of cognitive science, psycho and neuro-linguistics.

When it comes to explanatory influence, there is I contend, a clear asymmetry between the two interfaces classically referred to in minimalist theorizing. The externalization of the system is a highly variable and contingent, and known to preserve functionality across the modality of auditory vs. gestural sign. Externalization per se is arguably one of the design factors of language, but the exact mode of externalization demonstrably not. On the other hand, the domain of generalized cognition that syntax is embedded within presumably does not and cannot vary from language to language. Our highly complex human systems of thought and categorization are the source domain for language, and language is an extension of our superior cognitive abilities in apprising, categorizing the world. The human mind is the crucible within which language evolved in the first place, and while one does not need to accept the idea that language `evolved for communication’ of our thoughts about the world, it certainly plays a symbiotic role in our ability to both cognize and represent our own thoughts to ourselves.

Importantly, human language is not just a big bag of conventionalized symbols triggered by episodic stimulus in the world, it differs from other living creatures’ signalling systems in a number of striking ways, and share properties with those systems in others. Arbitrariness of the sign, and externalization of signal can be found in systems throughout the natural world, from monkey calls to mating dances, from songbird tunes to pheromones. But these collections of arbitrary signs do not have a syntax, and they do not systematically require the detailed tracking of the perspective of other minds; the property of creative open ended composition of meaning is unique to humans (apparently).

It is important therefore to emphasize that mere sign-sign relationships (systematic combinatorics) is not sufficient to create the natural language `magic’. Birdsong has been shown to have some form of syntax, in the sense of brute combinatorics, but does not have semanticity. By this I mean that the units that combine and relate to each other in systematic ways do not correspond to meaning units that also undergo composition in parallel. It is the combination of syntax and semanticity that creates open ended meaning composition which is the core innovation of the human species (cf. also Miyagawa et al. 2014 ). The recursive composition process that syntax affords is the feature that delivers creative meaning composition that can be coded and decoded reliably by human minds.

So in terms of deep properties of natural languages (NLs), I would argue that the interface with the internal systems of cognition is not an interface that is completely parallel to the interface with sound perception and production. The former is the interface with the source domain for the content being represented, while the latter is the contingent mode of externalization. Intuitively, the point of language is not to produce sound (this, for example, might be point of music), the point of language is the expression of internal patterns of thought (whether for oneself or others).

If we are to probe the question of what particular aspects of the mature working grammar emerge based on independent features of human thought, then we need to ask a different kind of question about the nature of the representations that the mature syntactic competence traffics in. Broadly speaking, we could characterize these questions as follows:

A. Are there mutually determining relations in the way syntax maps to thought, either with respect to primitive categories or the relations between them?

B. Are these hardwired or developmentally guided? What is the scope of variation in what is learnable in this domain?

C. To what extent do the features and categories reified in the syntax feed back into general cognition and start having an effect on our abilities to think abstractly and creatively?

These questions are at the heart of what I take to be the field of syntax-semantics within a specifically minimalist agenda, and of morphosemantics too, if morphology is a kind of syntax. Crucially, since the minimalist programme is stated in terms of the mind/brain of the individual, the core questions here refer to internal representations of meaning. So the Syn-Sem interface here is actually a somewhat different research programme from the one formal semantics is classically engaged, and is based on an understanding of what I-Semantics looks like. Unlike the intermediate representations of E-semantics, there is a right or wrong of the matter when it comes to postulating I-semantic representations. This already brings its own methodological differences and challenges. More of which in my next post.

The question of semantics and its relationship to syntax is often confusingly entangled with questions concerning the autonomy of syntax, controversies around the nature and scope of Minimalism’s universalist claims, and further confounded by the many different implicit definitions of what semantics itself is taken to be.

At a descriptive level, language systems are complex, involving both form-based facts as well as generalizations related to effects on truth conditions. The overall system is clearly modular. Some effects on truth conditions are not part of the syntax itself but of pragmatics and inferential processes. Some effects on acoustics/articulation have to do with phonetic implementational algorithms that are not part of the narrow linguistic computation. At least when it comes to inputs and outputs anchored in the external world, we can pose objectively clear questions. For sound, the externalized and objective acoustic reality is measurable and quantifiable, and we can ask questions about the gap between that measurable output/stimulus and the internal representations that generate it (or perceive it). Similarly, meanings of sentences can be anchored through detailed descriptions of the external realities that make them true. We can thus ask about the relationship (gap) between the linguistic representations produced/comprehended and the truth conditions associated with them in context. These interface questions are important questions of synchronic grammatical description. So for meaning specifically we can ask what part of the mature speaker’s behaviour is controlled directly by their linguistic competence, and what part is controlled by their pragmatic competence. Our pragmatic competence of course is finely tuned to subserve language, but it also arguably operates in non linguistic domains.

But, I would argue that when it comes to the nature of the syntax-semantics interface. specifically, it is much less clear what the question is about, since linguists seem to disagree on what is meant by these labels, and they do not have neutral architecture/ideology independent definitions.

In a mature grammar, syntax as a computational system operates with its own relations and primitives which, by hypothesis, are not reducible to either semantics or phonology. This is the gist of the autonomy of syntax proposal put forward by Chomsky as early as 1957 in Syntactic Structures, and maintained explicitly by Chomsky in 1982. Adger ( 2017) gives the following articulation of the principle:

“. . . syntax as a computational system that interfaces with both semantics and phonology but whose functioning (that is the computations that are allowed by the system) is not affected by factors external to it“. Adger (2017)

In terms of the functioning of a particular grammar as part of a person´s synchronic competence, it is relevant to ask what sort of system the combinatorial engine is, and what determines its functioning. For example, if we consider the interface with systems of externalization such as phonology, most would agree that syntax does not have rules like the following (1)

(1) Allow a syntactic unit which begins with a plosive to syntactically bind a syntactic unit that begins with a liquid, but not vice versa.

It appears that a kind of modularity that operates between the phonological system and its categories and relationships, and the syntactic system proper, which operates in a way that is blind to the difference between a stop and a liquid. This makes sense given that modes of externalization can vary, and that externalizing via sign for example gives rise to human language, with all the distinctive characteristics we are trying to understand, just the same as vocalized externalizations. In each case, the internal representations have to be systematically externalized for uptake by others, and then decoded and interpreted reliably by their own minds.

Do we have a parallel to (1) when it comes to semantics? Indeed, nobody believes that there should be a rule of syntax that distinguishes directly in terms of referential content.

(2) Perform syntactic operation R on a syntactic unit A, if it refers to a mammal in an actual context of use.

This would cause syntax to treat some of its nominal units the same way based on whether they ended up referring to a mammal in a particular utterance. It would mean that the name Fido would have to behave differently syntactically if it were the name for a dog, or for my pet lizard. And it would force my dogto pattern with one of those Fidos and not with my lizard . It seems bizarre to think of a language with such syntactic rules.

This seems fully parallel to the non-rule of syntax making reference to phonological segments, and at first blush argues for a parallel autonomy of syntax from semantics. But is this the right, or equivalent, analogy? Phonological features are, after all, abstract mentally represented generalizations over auditory percepts, not acoustic reference itself. It is important to note then, that there are cases of syntactic rules that appear to make reference to `semantic’, or interpretable features. For example in languages that care about animacy for subject selection. Even in those cases though, language seems to care about the abstract syntactic category rather than the actual denotational facts, or even conceptual category facts about what the culture conceives of as `animate’ or not, so that mismatches are found between abstract linguistic classification and cognitive judgements.

Consider also, as in (3) here, a toy rule which needs to make reference to a [+Q] feature, correlated with being interpreted as a question.

(3) Move the syntactic element in T to C if the latter bears the feature [+Q]

But, this is not a counterexample to the autonomy of syntax from semantics because [+Q] is a syntactic feature by hypothesis. And though it might systematically be translated via an abstract interrogative semantic representation, it is not exceptionlessly correlated with an actual questioning speech act in practice.

Pragmatics may intervene as in (4-a) so that the outcome of [+Q] being present in the structure is not actually a request for information; conversely, the request for information in (4-b) does not in fact require the syntactic feature [+Q] that is responsible for the movement of an overt tensed element past the subject. (Although distinctive intonation may be present, as is well known, this is actually dissociable from overt question-movement).

(4) a. Is the pope catholic?

b. You broke my favourite vase?!

It is widely acknowledged that the actual form of utterances radically underdetermines their truth conditions. Aspects of context, anchoring of indexical elements, resolution of anaphoric dependencies and conversational implicatures triggered by the particular the discourse context are all required before truth conditions can be specified. Nevertheless, the syntactic representation does provide a foundational skeleton of meaning contribution which is an important ingredient of the concrete meaning intended and apprised in context.

Because of their inter-subjectivity, truth conditions have seemed a convenient, plausible (and indeed necessary) way of grounding discussions about what sentences of a natural language `mean'. There is a well established use of the term semantics to pick out facts of reference and truth in an external, non-linguistic domain. I will refer to this use of the term semantics as E-semantics. The use of the feature [+Q] in syntactic theorizing can never be replaced by E-semantic facts concerning requests for information. Obviously. But let us imagine that we can isolate a contribution to truthmaking that was always the `translation’ of the syntactic feature [+Q], as a component of some kind of intermediate semantic representation. Syntax would still be autonomous in the sense that it manipulates syntax-specific units of syntactic representation, it is the translation algorithm that maps these syntactically active features onto something regular in a corresponding (intermediate, and still language-specific) semantic representation.

Very many features standardly assumed in syntactic representations are suggestively labeled with words that gesture towards a kind of interpretation (interpretable features), but which are fully paid up syntactic club members in practice. Their status as [+interpretable] only refers to the fact that such a feature is in the domain of the translation function from the syntactic representation to whatever representational form the underspecified linguistic `meaning’ occurs in.

I suspect that most working linguists and semanticists believe that there is an independent semantic representation which operates with different primes and primitives from syntax, and the question of the syntax-semantics interface is a question of how the primes and primitives of the one kind of representation translate into the other. The challenge here is to understand the systematicity of that relationship, which somehow must hold, if children are to acquire the ability to creatively generate meanings of complex utterances from component parts. Given the role of context, no direct mapping between form and E-semantics is possible. But if there were an intermediate semantic representation generated by a systematic translation algorithm, then the gap to E-semantics could be filled in by studying the systematic relationship between that semantic representation and actual truth conditions. (This latter is what I take to be the traditional understanding of the field of pragmatics.)

One standard understanding of the syntax-semantics interface, then, is the study of the translation algorithm that operates between syntactic and this intermediate semantic modes of representation; pragmatics is the study of the inferential processes that fill the gaps between the intermediate semantic representation and the fully precisified representations that can be paired up with truthmakers.

What is the status of this intermediate semantic representation itself? If it is on the language side of things, then doesn’t that mean that the strong minimalist thesis is false?. If it is on the non-language side of things, then its abstractness, its variability from language to language (which I think is an undeniable empirical fact), and its mismatch with non-linguistic categories of meaning are difficult to account for.

Posing the question in this way however, would be in my opinion, a misapplication of the strong minimalist thesis. For the strong minimalist thesis is not intended to hold at the level of descriptive modularity in the mature speaker’s system of competence, but rather refers to the role that other properties of mind/brain play in how the final system emerges. It refers to how we understand the initial state of the language faculty, and what (if anything) we need to put in there to ensure that the language systems we describe have the properties that they do. The strong minimalist thesis is about explanatory modularity, about what is present innately in the human brain that makes language possible. It differs from earlier incarnations of Chomskian writings in postulating a larger role for independent properties of mind/brain (which also could be unique to us, and also possibly innate). The minimalist programme says it is more `minimal’ to explain language properties through things we have to assume anyway about human minds, than to invoke language specific devices.

Barbara Partee, in one of her recent papers on the history of formal semantics within the generative paradigm makes the same point, in attempting to explain why Chomsky’s own attitude towards formal semantics (of e.g. the Montagovian type) has been often quite ambivalent:

“. . . it has seemed to me that it was partly a reaction to a perceived attack on the autonomy of syntax, even though syntax is descriptively autonomous in Montague grammar. But syntax is not explanatorily autonomous in Montague grammar, or in any formal semantics, and I do not see any rational basis for believing that it should be. The child learns syntax and semantics simultaneously, with undoubtedly a great deal of `innate knowledge’ guiding the acquisition of both (Partee 2014 pg 9)”

If we take the strong minimalist thesis in the sense of explanatory autonomy, then it is still perfectly consistent with that thesis to assume a language specific system of abstract semantic representations, correlated with syntactic forms, and not identical to the way non-linguistic cognition is structured, but in turn interacting with it. The mature system, the semantic representations and their internal vocabulary are in some sense hybrid representations that are not the same as that provided by cognition more generally, because they have been constructed over the course of acquisition to interface with syntax in order to solve the particular problem of codification and creativity. To quote Partee (2014) again,

“. . . syntax should provide the relevant `part – whole’ structure for compositionality to work.”

One can work on problems of the syn-sem interface in this sense without that being inconsistent or contradictory with the strong minimalist thesis. That is because it is perfectly possible that the final complexity of the system is emergent based on some rather simple abstract initial ingredients, only a small part of which is unique to language itself. The way in which language is set up is heavily determined by the need to interact with non-linguistic cognition (among other things), and this constrains the concrete systems that emerge in practice. Still, I would say that someone working on the syntax-semantics interface in the sense of constructing a computational theory of the mature grammar is not actually pursuing the minimalist agenda directly, even though they might be sympathetic to it or have it in the back of their mind as an important project. The descriptive patterns of actual syntaxes and generalizations about how they map to our agreed format for semantic representations are however surely part of the data that will be important empirical ground for other families of theories exploring the question of the explanatory role of cognition in constraining the general form of natural languages. So the existence of the subfield of research exploring the syntax-semantics interface is not a threat to either the autonomy of syntax , or in contradiction to the strong minimalist agenda.

In Part III, I will explore a somewhat different approach to the syntax-semantics interface, which is more directly engaged with the other project of exploring the questions of explanatory modularity at the heart of the strong minimalist agenda.

The central insight of the Minimalist Programme is that we should seek to pursue a model of grammar where the properties and core operations are pared down to the minimum necessary ingredients that can interface successfully with the non-linguistic aspects of the human mind/brain, while fulfllling the complex symbolic functions of natural language (Chomsky 1995b). Under this view, there is only one system, and the only interfaces are with non-linguistic cognitive systems. Moreover, the core computation that makes language possible (let us call this `syntax´) is highly constrained by virtue of the fact that it must interface with these independent properties of mind/brain, i.e. general cognition and things like long term and working memory. The minimalist conjecture is that these constraints imposed by the interface go very far in determining the specific form of language, and that the part of the system that is independent, and specific to language is actually quite minimal, possibly restricted just to the bare fact of (automatized) generalized recursion, or Merge (Hauser et al. 2002).

Part of what is obscure in discussions of the role of the interface is that the other modules of mind/brain can potentially constrain the syntactic computation at a number of different relevant stages, phylogenetically, ontogenetically and as a continually accessed component of the synchronic system. Arguably, all of these points of influence from general cognition exist, and help to determine the final state of any individual human´s grammar. We can engage in questions of the evolution of language independently of study of the final state of the grammar attained by any individual human. And the final grammar that belongs to a human being will certainly have more in it descriptively speaking, than recursion, even if the minimalist conjecture is correct that it is the only language specific hardwired universal. Putting aside the memorized lexicon for the moment, the actual system that a user of a natural language operates with in real time involves categorization (labels), category relations and linearization algorithms that become part of procedural memory. All of the specifics of these components can be language particular, but all languages seem to need these components, including the rich conceptual lexicon.

So the salient question then is, how much of this necessary detail on top of Merge is emergent, based on the constraints imposed by general cognition and other third factor considerations? The answer can only be arrived at by detailed theoretical analyses of actual natural languages, searching for generalizations which will ultimately allow us to explain some of the details via the properties of other modules of mind/brain, even when the incarnation of those effects within the language system is special to that local instantiation. To thoughtfully pursue a minimalist agenda, therefore, in addition to the description of languages and the building of generative theories, we need to collaborate with or at least keep track of research on the properties of the other systems of mind/brain that are implicated in the instantiation of language. There are many such potentially relevant cognitive building blocks that could have a constraining effect on the emergent system: working memory, conceptual categorization, access to long term memory, associative mechanisms, timing mechanisms, and cross referencing and control in parallel processing across submodules etc. to name just a few that spring to mind.

These are all things about human mind/brain that we now understand a lot better than we did even 20 years ago, and which exist independent of language but which, crucially, are harnessed by it. In fact, one can read the minimalist agenda as an invitation to aggressively pursue such sources as explanations for the details of syntactic description. In other words, the way the human mind works can reliably (and even universally) force a certain kind of syntactic system to emerge, without any of the cognitive precursors being specific to language itself, and without compromising the implementational autonomy of the mature system. The proposal that Merge, or more generally recursion, is specific to language, is actually a departure from a fully minimalist thesis, which would seek all the ingredients elsewhere, and seek the uniqueness of language in the coupling, or automatization or certain human innate cognitive abilities.

In this context, what do we mean by the `syntax-semantics interface’, and is it consistent with a minimalist agenda? In the next few blog entries, I will explore a series of themes related to this question with the ultimate goal of carving out a domain of inquiry which I think is vitally important and interesting, but which has so far, I think, been underexplored and often misunderstood.

My summer vacation. 4 weeks. 11 books. 8 countries (all imagined, one invented).

The most remarkable thing I read this year was the one I started before lockdown began, Györgi  Buzsáki´s  The Brain from Inside Out. In this often highly complex but compulsively readable book, Buzsáki argues for replacing the pervasive conception of the brain as a receptive organ (generating output reactions from sensory input striking it from external reality)  with a radical rethinking of it as a proactive  one:  an obsessive poker and prodder of reality, actively seeking to latch on to and resonate with the outside world. He supports this reconception with evidence from his long experience as a neuroscientist and shows how it utterly reconfigures they way we ask questions about cognition. Specifically, the internal physical properties of the brain and its neural architecture form the framing structure for cognitive experience, it is not an empty vessel waiting to be filled with statistically regular patterns from the world.  I am not exaggerating when I say that reading this book rocked my world and I have spent the rest of the year so far turning over old and new facts from linguistic processing from this perspective.    Of course, we all as humans have a sense of `the internal world´, our internal lives that construct an internal narrative and emotional landscape through which we face the world of social interactions. It is important to understand that this is not directly what Buzsáki is talking about. His revolutionary thought is about the level of neuronal signalling itself, a microneuronal analogue of an internal language syntax looking for a semantics. The brain is born singing and breathing,  and bounces itself off the world in order to know itself and its place within that world.

The rest of my summer reading was about feeding my inner world at the more macro level, the way the idea is more commonly understood.    In the first weeks of lockdown way up in the far north, I picked up Ron Jacobsen´s Unseen, translated from the Norwegian, a tough and elegaeic novel about a family unit eking out an existence in an unforgiving Norwegian island landscape. In the face of the unrelenting hardness of life, little is actually said, but the inner lives of the protagonists twist and roil under the surface.  The young female growing up in this bleak and choice-less environment grows into her role and accepts it; it felt to me like how it might feel to train your feet to be small through years of binding, or gradually growing into the cliff face you are strapped to over the course of time, but who is to say it is not in its own way a kind of freedom?  The young female of Unseen enacts no permanent rebellions, but the young female protagonist of Educated  by Tara Westover does. Her growing environment in rural Utah is presided over by a genuine psychotic father and completely paranoid and insane world view. She is actively prevented from learning or knowing anything about  the outside world and this autobiographical tale is the story of how she escapes. If you don´t know that you don´t know, how do you  learn to know that, so that you can seek genuine outside knowledge? How do you break out of the cycle when your upbringing has made it impossible to know or trust what is Real?  A quick and compelling read (not great writing, but curiosity propels one through to the end).  The fascination in watching this journey is made more relevant by the dark fear I have that there is a whole swathe of Trump supporters living under similar conditions of constructed reality, and for whom there is probably no redemption story possible. It is hard to escape the loops of disinformation if your whole belief system is predicated on Faith in the Unreliability of Other Voices.

Other Voices!   In the huge hurt and upheaval around 2020s Black Lives Matter movement, I threw myself into Colson Whitehead´s The Underground Railway,  a visceral and moving tale of a woman born a slave in the deep south, and her efforts to flee through the underground railway linking secret station locations across the South, built by slaves for helping other slaves escape. I will not tell you if she escapes because you should read it. In some ways, no, there is still no escape. Black lives in today´s America still bear the scars and systemic effects of this history. The details of this history should be taught unvarnished in schools, but also This, this beautifully written novel will clutch at your heart.

N.K. Jemisin author of The Obelisk Gate  is the first black female writer to win the Hugo Award for Science Fiction fantasy and this is the third book of the trilogy (I read the first two last year).  I don´t normally read much fantasy (I usually lose patience with the writing).  But this is a great, richly constructed reality with lots of prickly and violent and difficult female characters. And in this world, the very earth is unreliable, it seethes and boils with danger to the planet, and there are some of the human race who can tap into and control its heavings.  They are needed for the rest of the human race to survive,  but they also inspire fear and disgust.  So Yes, there´s race politics woven in here too,  if you are willing to look for it, imaginatively transformed within this fantasy world.

Next, I read Ingrid Persaud´s Love After Love which is set in Trinidad and written by one of my contemporaries (yes, and friend! ) so I was curious to see how she would represent our shared experienced world of  Trinidad in literary form. The challenge she pulls off here is in capturing the music and rhythms of Trinidadian speech, while still making it accessible to a non Trinidadian audience.  One of the central characters is a Difficult Woman too, something that was turning out to be a bit of a theme for me this summer.  But unlike the central male character of the book who came off the pages very vividly (a moving portrayal of a gay man in a profoundly homophobic culture), I was a bit disappointed in my central female here—– it seemed to me she teetered on the verge of self actuation throughout the book but never quite got there.

We watched Mrs America in the evenings while on summer vacation (brilliant!), and I was inspired to read Helen Lewis´ s  Difficult Women. In 11 Fights  next,  which is however more focused on the UK than the US.   This was a great window into a lot of history that I had sometimes heard of but knew very little about. Lewis doesn´t judge, or woke-wash, and these women come alive in all their idiosyncrasy and ornery-ness.  Rereading about the fight for female suffrage in Britain and the shocking misbehaviour of those early women revolutionaries, I am reminded that change in power structures doesn´t come about from asking nicely……

That led me to the next ornery woman on my journey, served up by Olga Tokarczuk , who won the Nobel prize for literature for 2018. I had already read Flights,  which I loved, so  I treated myself to  Drive Your Plow Over the Bones of the Dead. The protagonist is definitely a Difficult Woman. I loved being transported into her reality and watching the murder mystery evolve. The men in the hunters club  in a remote Polish village near the Czech border start being killed off mysteriously and our old lady is convinced it is the animals who have turned on them and are subtly orchestrating their demise. The writing here is a joy!

 I moved from here to an older classic, which I had somehow never read: Carson McCullers The Heart is a Lonely Hunter written, astoundingly,  when she was only 23. This had everything—- race, socialism, gender, sexuality and a quirky authorial voice always straining to get at the heart of things. The young female character, Mick, talks at one point about the Inside Room and the Outside Room of her own consciousness in a way that resonated, and took me back full circle to Buzsáki. Towards the end of the novel, Mick takes a job in a store to help her family make ends meet and has to wear a skirt and stockings. Her life becomes so busy that the Outside Room begins to elbow out the Inside Room (where her Beethoven symphony and conversations with Mr Singer reside). I realized too that even for privileged middle aged academics,  sometimes things get so hectic that the Inside Room starts to gather dust.  Fortunately, lockdown and the derailing of routines can I the best case lead to release from small tyrannies.  Still, it has not been so easy to retreat:  The Outside world has been staging its own can´t-look-away spectacle, raging around us in 2020 in the form of pandemia, BLM protests, and Free Speech Wars.   Binge reading with no external purpose for 2 weeks  in July has been my way of getting back into my own Inner Room, that sustaining and reinvigorating place, like the intellectual equivalent of nightly sleeping and dreaming after the requirements of the day.  My pile of books was cobbled together through accidents and serendipities—- I have no idea why McCullers got on the list but boy that book kicked ass. I won´t give away any more of the plot by explaining why.

But oh no, it´s nearly over!  Work was set to begin again, so I took one final plunge,  journeying imaginatively to the far East to read Min Jin  Lee´s  Pachinko,  a tale of Korean immigrants in Japan spanning the 20th century (continuing my themes of racism and otherness and difficult women).  Finally,  Cixin Liu´s  Three Body Problem,  a science fiction tour de force translated from Chinese.  Like N.K. Jemisin, this is all Difficult Woman Gets Pissed Off and Decides to Destroy the World,   but now that plotline is woven together with particle physics, online gaming, philosophical abysses exploring the relationship between data, theory building, prediction and faith in science.  Set this all against a background of the Cultural Revolution and you have enough food for the journey right back into the Outside room of doing science and coping with dystopian reality that is our post truth era.

Sigh. Time to wake up.

 

 

 

I have taken a break from blog posting recently because I tend to post when there is a conference or workshop or event that provokes excitement and self expression. Lately, with the corona crisis there have been no workshops to attend in person, and there is absolutely no point blogging a conference that is digitally available to all anyway!

I also have to confess that I joined twitter last year (only academic followees) and have been down many rabbit holes because of that which may have used up some of my time (lets be honest. most of my waking spare moments).

I am posting now because the Pinker letter thing has exploded and it turns out sometimes it is best to say your thing calmly, after editing, than to fire off  a series of tweets or respond in a piecemeal way to the facebook posts of others.  I am not a member of the LSA, and I think in many ways, the USA is  a social and even academic microcosm unto itself that requires subtlety and understanding and `on-the-ground’-ness to negotiate properly.  I do not think that my voice is the democratically relevant one in this fight, and my US academic colleagues have the responsibility and the burden here.   Here in Europe,  things are different in ways that vary massively from country to country, but it is also true that because of the influence of the US (for better or for worse) in our sociopolitical world and the world of linguistic academia, we are all in some sense affected and involved. So here is my two cents.

The issue in this particular case  boils down to democratic representation. Remember that?  That was the thing we had in politics before global capital and vested interests gutted our choices and hollowed out the heart of democracy. It was what we had before it became impossible to get simple facts out into the common public discourse, or to deny obvious falsehoods,  because the media are controlled and turned into instruments of tribalism with no accountability.   But still, democratic representation still exists right? On the lower and less policitized levels of nerdy groups and clubs, or even mid level academic societies, where we can decide who our steering committees are, and who we trust to make organizational decisions etc. etc.  Right?

I waver between being  despondent about the state of sociopolitics today and recklessly revolutionarily hopeful.  I think that we have been crushed by the neoliberal status quo for the past 50 years or so, so that often people can no longer  muster the social outrage appropriate to the kinds of inroads in opportunities and equality that have be meted out to us under the flag of global capitalism, to which there are genuine and viable alternatives. The recent debacle of the Trump presidency and its latest corona mismanagement, and the phenomenal and  diverse upsurge of political activism in wake of the George Floyd murder means that progressive ideas (especially among the young) are on the rise. But let there be no mistake about it. This has come despite the consistent and powerful forces of the political and financial elites who have thrown their whole might behind silencing these voices. They talk about cancel culture, they talk about fake news, they talk about conspiracies, and they consistently accuse the left of fascism and intolerance and repressing free speech. But this is their playbook, not ours. And the way public discourse works, it is IMPOSSIBLE to prove even that climate change is real, let alone that the current political system is the one that has silenced and is still silencing dissent.  But ask anyone of colour, anyone who supports Palestinian rights, anyone who believes in socialism, or who believes in disinvesting from fossil fuels completely, or indigenous rights (the list could go on) and look at how hard it is to get that message out. And compare that to the dominant voices in the mainstream media.  I believe in free speech, like the majority progressives. And in accepting the consequences of one’s positions, and in open discourse and disagreement.  Famous heavyweight intellectuals waving the free speech flag to censure non renowned linguists for writing a 4 page letter to a nerdy academic institution frankly dismays me.

The Pinker battle is precisely about who gets to speak for us (as an academic discipline) in the mainstream media. It is about democratic representation. We say as a field in other contexts that we need to be better at communicating and being involved in public discourse.  Language is hugely relevant to our lives as social and political beings, so we are important here.  SP has a loud and dominant voice in the current climate anyway, because of his many books and contacts, but he has also been granted a platform by the leading academic body of the field in the US (the LSA)  to speak for us on issues of general linguistics (one of two names on the General list for media contact!).   So my question is,  suppose the next generation of young linguists (500 of them, or so) who have been engaging in public discourse as part of their beginning careers, suppose they feel that SP does not speak for them ? Suppose they are constantly dismayed, ashamed and feel their intellectual positions undermined by the things he says on twitter, or in the mainstream media?   What do they, as members of an academic institution do?  What do they do?  What is allowed to them in this context?    Mind you, this is not to change the world, or eradicate systemic racism or save the planet or chip away at the smug unassailability of global capitalist logic—-its  just one simple thing:  to figure out a way so that their small corner of academic where they live and work is sound and good, and whose intellectual integrity they can be proud of and get behind.   They write a letter to their executive council saying, hey guys,  a lot of us think this dude does not actually represent our ideals, can we take him off that damn media list now already.

Now even if you don’t at all get it that SP is this bad,  (but read the articles and quotes over an extended period of time and make your mind up for yourself), at least recognise that a large number of young and passionate linguists whose work and integrity you respect felt moved to sign that letter, even though some of the argumentation might not have been ideal, and even though they were torn about academic freedom issues.  They signed it and they are not hooligans with baseball bats destroying all that is good and sacred. They were trying to have a say.  We should trust them and look carefully at that media list again, and try to figure out ways of having a system in place so that democratic representation is respected.  After all, we want our institutions to reflect the reviews of its members, and they deserve to have an institution that represents them.

The letter writers are not cancel culture vultures seeking to stifle free speech as a part of a progressive agenda gone mad (which is how I am reading some of the reactions). They were just trying to have a say.  (How hard it is to break through the clamour of the already established voices and privileged positions!  The problem with the world today is not that neoliberalism is having a problem getting its message out.)

The future of our field, and indeed of our planet is with our young ones.  We should listen to them, and trust that they do have our values of free speech and academic integrity. They are changing the field towards better support for:  female linguists, early career linguists, linguists of colour, LGBTQ non-binary linguists, indigenous language issues etc. etc, in ways which are driven by them not the old guard or us middle aged fuddy duddies.  Long overdue. We should listen to them if we are the ones who happen to have the power,  not completely onesidedly,  but as part of a respectful open discussion.  I for one believe that if we create an academic community that meets their standards, it will be a good and fine thing.  The LSA should take this letter as an opportunity to make some changes, and put in place some mechanisms for change and updating more generally.

 

I am hereby filing my official report on the second incarnation of the OASIS meeting, OASIS 2 ( Programme), with its satellite workshop on correlating possibilities (programme) .

I am on the record as being a cheerleader for crossdisciplinary research. One personal and perhaps not that interesting reason for that is that every time I stretch outside my comfort zone I learn something which excites, or titillates and which allows me new insights for my own local research project. But the more urgent reason in the case of semantics and ontology is that I believe there are scientific questions that simply cannot be answered satisfactorily without linguists and neuropsychologists comparing results and joining forces. The scientific question I am referring to of course is the question of creative meaning composition, and how human minds pull off the amazing trick of using symbols to create meaning/meanings in an open-ended way.

Currently it also seems to me, and to the members of the OASIS network, that there is simply not enough crosstalk between theorists of meaning  (semanticists) and the fields of syntax on the one hand and psycholinguistics/neurolinguistics  on the other. We care about different data and we ask different questions about it.  So is this unavoidable, and if not, what can a conference like OASIS do to start to redress the situation?  More on that at the end of this blog entry.

Here are my thoughts about what went down this week  and what I learned, as a way of sharing information for all of you out there who were not able to be there. But bear in mind, that, as usual for this blog, this is a personal and highly opinionated riff on the contents of the conference mediated by my own idiosyncratic interests.

Is Syntax Real?

David Adger (Queen Mary University of London) kicked off the first day with a class/tutorial on What Syntacticians Think  for the non-syntacticians in the room.  Adger is an excellent speaker and is extremely engaging when presenting syntax for the non specialist. Adger’s remit was to convince the crowd that syntax is real, and that once the full range of data is acknowledged, one simply cannot make do with just sequence, memorized chunks, and transitional probabilities.  But here  our  failure as OASIS became immediately obvious—- nearly everyone in the room was already familiar with syntax,  and with Adger’s take on it. (We had utterly failed to reach out beyond our discipline. So poor Adger instead had to talk to a basically sympathetic, already clued-in crowd.)  We tried to draw out some more contentious issues in the question period. Here’s one that provoked an interesting flurry: He was at pains to emphasize that the syntacticians `derivational talk’ refers to issues of logical dependency and is not intended to refer to real time operations. The statements describing the grammar are entirely logically  separable from the heuristics that are employed to decode incoming speech and code our expressive output using that grammar.  Ramchand intervened to remark that that was all very well and good, but then you still owe us a theory of the relationship between the two, otherwise you are in a prediction vacuum. One could argue that EVERY SINGLE piece of data that you might use as your evidence for theory building, is a piece of psycholinguistic behaviour, either of production or comprehension and that since there is no way of accessing competence directly except through performance manifestations in a variety of task sensitive contexts, we cannot proceed without a theory of how processing relates to the grammar, if we are to have any data at all.  Ramchand then goaded the speaker by saying  that many generativists claim that the grammar does not make direct claims about real brains in real time, but then in practice, they lapse,  and use precisely these kinds of considerations in justifying the shape of their theories.   A lively argument ensued.  Which was inconclusive and had to be forcibly broken up by the chair of the session, much to the disappointment of the audience.

A Living  Inter-disciple

The second talk/class was by Naama Friedmann  who is a self-described neuropsychologist of language from Tel Aviv University.  Friedmann had a rapt and captive audience of linguists who told us all the stuff we wanted to hear. For example, her studies on Specific Language Impairment (SLI) in Hebrew and Palestinian Arabic show a very particular profile of impairment for a sizeable percentage of those officially diagnosed with SLI (language impairment with otherwise normal intelligence).  It turns out that they simply cannot handle sentences where the relative linearization of two DPs violates the standard positions expected from thematic role under normal word order conditions. In syntax jargon, they can’t do crossing movement and sentences like (1) and (2) below are simply impossible for them to comprehend and produce (and resumptive pronouns don’t help, and neither does prosody).

(1) Which girl did grandma kiss?

(2) The girl who grandma kissed is smiling.

It can also be shown that it is not working memory that is the problem, since subject relatives with long displacement do not cause the same kind of problem.  The direction of the movement is also not relevant, since this work was also replicated for Israeli Sign Language where one of the movements was rightward.  SLI exists for signers as well, and displacement of this type is the one specific thing that they cannot do.  Wonderfully and heart-warmingly, Friedmann has developed teaching/intervention strategies that actually are successful and allow people with this diagnosis to achieve excellent functionality via explicit instruction. This is in and of itself an interesting fact, as well as being a useful one to know.

What causes SLI? Not surprisingly, it appears to have some genetic sources. But it is also environmentally triggered and can be affected by events early in life that have profound effects on neural developmental. Staggeringly,  due to a very specific  baby formula disaster, a group of hundreds of babies in Israel suffered severe B1/thiamine deficiency in their first year of life and were hospitalized, treated, and subsequently followed throughout their lives. Ninety-seven percent of this cohort subsequently developed SLI.   Thiamine plays a central role in cerebral metabolism and a structural role in membrane structure and synaptic transmission. It turns out that thiamine deficiency in the first year of life leads to this particular kind of syntactic SLI which persists into adulthood, even though thiamine levels were subsequently normalized. The 1^styear of life window was also implicated in their study of deaf children of hearing parents. The kids who received hearing aids within the first year of life went on to develop normal syntax, but those for whom the discovery of the hearing problem was delayed, also developed this particular form of syntactic SLI.  So, weirdly, for acquiring one’s first language, there are developmental events happening in the first year of life that are necessary conditions for acquiring the ability to acquire flexible linearization of certain complex syntactic representations. It is really interesting that the ability/deficiency is so specific and relates to complex syntactic processing and yet the seeds of it are necessarily sown early and affected by input as well as nutrition affecting brain development.

Here’s another cool thing.  Suppose you were interested in whether the tendency to give an exhaustive answer to a Wh-question, versus a `mention one’  answer to a Wh-question is a matter of semantics/linguistic form or  of pragmatics, you might set up an experiment to compare the responses of populations with documented deficits in theory of mind with populations that have no such deficit.  The experiment that Friedmann and colleagues did involved both simple Wh-questions and multiple Wh-questions, together with pictures. So imagine a picture of different people holding different fruits (and some holding non-fruits). You are given the following questions and have to provide answers:

(3) Simple: Who is holding a fruit?

(4) Multiple: Who is holding what?

It turns out that Theory of Mind impaired people do not  consistently give exhaustive answers to simple Wh-questions as in (3), unlike the controls who tend to do so. This suggests that the choice between giving an exhaustive answer or a mention one answer is pragmatic and involves reasoning about other people’s reasons for asking the question.  But here comes the really interesting part: when it comes to multiple Wh questions, everyone, even the Theory-of-Mind impaired participants, gave exhaustive answers for the multiple Wh-question. So there has to be something about the syntactic/semantic/linguistic form for this type of question that demands exhaustivity.

Great. Sorted.

Events, Event Types and Connecting Verb Names to the World

Sudha Arunachalam (New York University) reported on a series of experiments she has been doing on children’s acquisition of event concepts in the form of event nominals (where there is on the surface a mismatch between the nature of the referent and the canonical syntactic category that realises it).  It turns out that kids’  beliefs about the meaning of event concepts differ from adults in a number of systematic ways.  In an experiment in which George Giraffe wants to learn the word nap,  children were instructed to point to the picture that would help George learn the new word.  It turns out that children are perfectly cool with nouns like nap  or party denoting events, but they also seem to have a more rigid criterial association between the event and the prototypical objects that the event usually involves.  So in the case of nap children seemed to think that a nap was an event that required a bed, or it doesn’t count as a nap (although they were clear that a nap was not in fact a bed).   In general, I think that the  Gavagai problem for verbs is even more severe than it is for nouns. Are there cognitive tendencies for what situational properties are associated with the verb name kids are learning?  It seems like identity conditions for events are harder to spell out than those for objects. Could it be that kids systematically latch on to object participant properties as a way of classifying and categorizing events for the purposes of verb word learning?    Nicola Guardino  (ISTC – CNR) talked in a rich detailed way about how events are distinguished and classifie.   In truth, the world is extremely rich in the sensory detail it presents to us, and the common sense idea of an event in the world involves a host of contextual information in addition to  the core and non core participants invoked by the predicate. The interesting linguistic question here is about  names  of events and what they get associated with, in particular the  in which our attention is focused on certain aspects of the situation and not others for the purposes of categorization and naming.  According to Guarino, there are different levels of involvement of event participants and a separation between focused vs. backgrounded subparts of events,  but importantly there is a further separation between event and the context it is embedded in. Theorizing about  cognitive classification is surely relevant for understanding the process of concept acquisition, and also for understanding b the pattern of verb types and verb alternations that are possible when it comes to naming events in language.   This turns out to be relevant to the problem of how to define minimal situational exemplifiers for a particular proposition that can feed a coherent notion of content. Angelika Kratzer(UMass) participated via skype and launched the defence of her version of situational truth making in the face of Kit Fine’s  version of truthmaker semantics. The debate between the two is relevant to the discussion about event classification and naming.   The disagreement centres on whether it is possible to get a coherent notion of Contentfrom truthmaking, if one assumes that verification of a proposition by a situation is generally taken to be persistent — if any superset of a verifier is also a verifier. In other words, if a situation verifies a proposition then any super-situation of that situation verifies that proposition. This means that if you want to understand the Content  of John swimming one kilometre, you need something better than just the whole world which has this situation as a subpart. Intuitively, you want something more minimal, something that is justthe situation that verifies John’s swimming one kilometre without any extraneous unnecessary bits.  And here is where the disagreement comes in. Kit Fine thinks that if you start with a notion of verification that is inherited upwards then the job of defining minimal verification in this sense is just too hard, and cannot be cashed out in terms of situational mereologies. He thus proposes to make verification exact (so that it doesn’t persist for supersets), and makes the fact of some situation being a truthmaker or falsitymaker for a proposition a primitive. Kratzer disagrees. She argues that her version of situational truthmaking does not require us to throw out forty years of formal semantic research on truth conditions. She just has a strong motivation for making the definition of minimal situation work. A large part of her talk consisted of working through a number of examples where the definition of minimal situation runs into problems and motivates some ways of getting around these problems. Kratzer’s slides and talks on this topic are available for viewing, since this material also formed part of the prestigious Leverhulme lectures that she gave earlier this year (Leverhulme lectures 2019).   I asked Guarino what he thought about the debate because it seemed to me he was thinking precisely about what parts of situations were extraneous to verification and what parts were essential.  Guarino was at first uncertain that he had anything to stake in an argument between a philosopher and a linguistic semanticist, and insisted that to the extent that he was a philosopher he was actually interested in metaphysics rather than in ontologies motivated by natural language. But in the end, he agreed that language mattered and that people talked about things at a particular granularity which was the granularity that he was interested in ultimately. Apparently this is called  descriptive metaphysics.  My question to myself was whether descriptive metaphysics might help Kratzer to define the minimal situation, without the paradoxes that mereologies over situational particulars give rise to. Or maybe descriptive metaphysics is relevant to unpacking Kit Fine’s primitive of exact verification.  Guarino claimed not to have an axe to grind there, but it was fascinating as an example of people talking about the same thing who are clearly also not talking about the same thing.  I was also the only person in the room who thought that Kratzer’s and Guarino’s talks were related (Except maybe for Orin Percus).

Louise McNally (University of Pompeu Fabra) presented her joint work with Scott Grimm on ing-nominals. One standard story in the literature is that while (5), the Poss-ofstructure, denotes a concrete event, (6), the Acc-ing nominal,  denotes a fact or a proposition.

(5) Al’s raking of the leaves/The raking of the leaves. (Events?)

(6) Raking the leaves/Al’s raking the leaves/Al raking the leaves. (Facts?)

The former are good with predicates like took place at x time, while the latter are not. The former do not support negation, while the latter do.  McNally and Grimm however argue that invoking a fact or propositional meaning for the forms in (6) is NOT correct. In fact, what McNally ends up proposing is that the of-ing nominals denote event tokens, while the accusative taking ones denote event types.   Basically, what McNally assumes about the semantics is that the ontology includes a type/token distinction for events, just as for entities. Nominal (of-taking) forms and verbal (acc-taking forms) both  lexically describe event types, but then grow to token interpretations in different ways: ing nouns through number morphology; ing VPs through temporal anchoring from a tensed verb.

The type vs. token distinction turns out to account for the data better than assuming that the forms in (6) denote propositions. Louise McNally is one of the few semanticists who is working on the idea of event types as primitive members of the semantic ontology. It’s really hard to spell out how this idea of ontologically primitive types  works when it comes to (i) compositional operations over types and (ii) the conversion of type level information into token level information. However, I am excited about this work because it is vitally necessary and  there isn’t anyone else really doing it. The problem of messy lexical content (polysemy, essential conceptual content etc.) is mostly ignored by formal semanticists.  And yet we know that memorized concepts are a core ingredient of the creative meaning generation capacity. Ignoring this messy part and how it gets deployed to create propositions that ultimately bear a truthmaking relationship with the world, is  abdicating our responsibilities for answering that Big Question. I think McNally’s work is an example of s the kind of work that the OASIS group is trying to support and create dialogue about.

Having said that, I think that ultimately we in the OASIS network have failed to be the place where the big interdisciplinary conversation is happening. What we just had was a wonderful `linguistics’ conference with lots of stimulation for us linguists from other fields. It is not clear to me that we have created a forum where linguistic expertise is being understood and internalized by psycholinguistics or neurolinguistics in building descriptively and explanatory models of meaning.  It is mostly our fault as a field.  When was the last time  we made choices within our analytic frameworks in order to line up  with the expertise and acquired knowledge from these other areas?   So OASIS 2 was  great (Thank you, Nantes team!),  but  I am not satisfied. I want more. When it comes to semantics there is a lot of hard work to be done and I really think that progress is not going to be made unless we get the big conversations going,  pooling our knowledge across disciplines.  I hear a rumour that the next OASIS meeting will take place in Tromsø, where we will have another shot at changing the world.  My fantasy title for that conference is Compositionality and the Brain  and I am already working on my fantasy football team for that event….