Computation in Language

For some time now I’ve been pursuing the idea that language is the basic locus of computation in the human mind and, correlatively, that it is grounded in processes that are non-computational in kind. This might not seem strange to a computational linguist, but other of course have other ideas. Many humanists are at best skeptical, if not horrified at the idea that the mind is computational in some way. And many cognitive scientists would like to think it’s computation all the way down to individual neurons. Humanistic skepticism is simply out of date while the cognitive scientists are over-eager, especially now that we have other ways of thinking about basic neural processes (e.g. complex dynamics).

More specifically, I identify computation with establishing the mapping between a language string and elements of meaning, whether in comprehension or production. The problem, of course, is that the elements of the string are all there together in one place, one after the other. But the elements of meaning are scattered all over the place in neuro-mental space. Here and there might be a pair of contiguous elements, but for the most part, not. We need a computational process to establish the coupling.

But why? Or, conversely, why doesn’t this problem exist for any other activity?

Questions questions questions.

Could mainland China force itself on Taiwan through military power? Perhaps not.

Tanner Greer, Taiwan Can Win a War With China, Foreign Policy, September 25, 2018:

China has already ratcheted up economic and diplomatic pressure on the island since the 2016 election of Tsai Ing-wen and the independence-friendly Democratic Progressive Party. Saber-rattling around the Taiwan Strait has been common. But China might not be able to deliver on its repeated threats. Despite the vast discrepancy in size between the two countries, there’s a real possibility that Taiwan could fight off a Chinese attack—even without direct aid from the United States.

Two recent studies, one by Michael Beckley, a political scientist at Tufts University, and the other by Ian Easton, a fellow at the Project 2049 Institute, in his book The Chinese Invasion Threat: Taiwan’s Defense and American Strategy in Asia, provide us with a clearer picture of what a war between Taiwan and the mainland might look like. Grounded in statistics, training manuals, and planning documents from the PLA itself, and informed by simulations and studies conducted by both the U.S. Defense Department and the Taiwanese Ministry of National Defense, this research presents a very different picture of a cross-strait conflict than that hawked by the party’s official announcements.

Chinese commanders fear they may be forced into armed contest with an enemy that is better trained, better motivated, and better prepared for the rigors of warfare than troops the PLA could throw against them. A cross-strait war looks far less like an inevitable victory for China than it does a staggeringly risky gamble.

UBS at sunrise

Notes toward a theory of the corpus, Part 2: Mind [#DH]

Way back at the end of September I posted the first of a two, possibly three, post series: Notes toward a theory of the corpus, Part 1: History. I figured the second post would follow within a couple days and perhaps a third a few days after that.

I got delayed, diverted by other matters. But here we are with the second post.

As I said at the beginning of that post, by corpus I mean a collection of texts. The texts can be of any kind, but I am interested in literature, so I’m interested in literary texts. What can we infer from a corpus of literary texts?

Then I was interested in history. This time around I’m interested in the mind. History and the mind, two different, but not unrelated, phenomena. The fact that a given corpus consists of text by many different authors is essential to making historical inferences. The different authors published at different times; we can use a corpus of those different texts to arrive at inferences about historical process. In that earlier post I used Mathew Jockers’ Macroanalysis as my example. He had a corpus of 3300 Anglophone novels from the 19th century.

In this post I’ll be using Matthew Gavin’s work on a passage from Paradise Lost [1]. He uses a corpus of 18,351 documents drawn from Early English Books Online and dating from 1649 to 1699 (which covers the years when Milton wrote and published his epic). I don’t know how many authors are included in the corpus, but it doesn’t matter. Gavin is interested in only one of those authors, John Milton. And while he says nothing about Milton’s mind, he writes only about his text, I will argue that he is, in fact, investigating Milton’s mind. But also the mind of any sympathetic reader of Paradise Lost.

How is that possible? On the contrary, how is it not possible?

The corpus in vector semantics

For reasons I’ll explain shortly, if only superficially, Gavin needed a large body of texts to create a vector semantics model he could use in investigating Milton. I don’t know how many words were in those 18,000+ texts, but if each text were, on average, 10,000 words long, that would be a total of 180 million words; 50K words each would yield a corpus of 900 million words. I figure we’re dealing with 100s of millions if not over a billion words or continuous text. If Milton had written a lot Gavin could have used a corpus consisting entirely of Milton’s own texts. Milton didn’t, so Gavin couldn’t.

However many authors are included in that corpus, each has their own mind. And, at the margin, their own idiolect as well. Still, they hold English as a common tongue; if it wasn’t pretty much the same for each of them it wouldn’t function as a medium for communication. As long as we remind ourselves of what we’re doing, we can treat the lot of them as one somewhat idealized corporate author of the texts in the corpus. It is the semantics of that author that Gavin is applying to Paradise Lost.

Given our corpus, how do we create a semantic model? At this point I am going to more or less assume that you’ve already been through a good explanation of how vector semantics works. But I’d just like to remind you of some aspects of the process.

The process is founded on the assumption that words that occur close together in texts share some aspect of meaning. How can we turn that insight into a usable model? We create a co-occurrence matrix.

Here I’m using an example from Magnus Sahlgren [2]. Consider this text from Wittgenstein, “Whereof one cannot speak thereof one must be silent.” It contains nine word tokens from eight word types, to use terms from logic (one appears twice). Now we need to define what we mean by “close together”. Given two words, are the considered to be close if they are, for example, 1) within they same 1000 word string, 2) immediately contiguous, or 3) something else? For this example let’s choose the second criterion.

Given that, we can produce the following table:

	whereof	one	cannot	speak	thereof	must	be	silent
whereof	0	1	0	0	0	0	0	0
one	1	0	1	0	1	1	0	0
cannot	0	1	0	1	0	0	0	0
speak	0	0	1	0	1	0	0	0
thereof	0	1	0	1	0	1	0	0
must	0	1	0	0	0	0	1	0
be	0	0	0	0	0	1	0	1
silent	0	0	0	0	0	0	1	0

It makes no difference whether we read by rows or columns as they are the same, but let’s read it by rows. If the word in a column is next to the target word we place a “1” in the column, otherwise “0”. So, whereof does not occur next to itself; a 0 goes in the first column. It does occur next to ˆ; a 1 goes in the next column. An so it goes for the rest of that row and for the rest of the rows. 

Note that as whereof is the first word in our text it can have only one word next to it. The same is true for silent, the last word. Since one occurs twice it is next to four other words. The remaining five tokens – cannot, speak, thereof, must, be – each have two neighbors, one before and one after. 

Each word is not associated with a string of eight numbers, which we can call a vector. We can now use those numbers to associate each word with a point in a space of eight dimensions, one for each word. In practice we wouldn’t bother with a corpus consisting of only one short text. But the principle remains the same given a corpus of 100s of millions of words constructed of tokens drawn from a population of, say, 100,000 types. Define what you mean by context and construct a co-occurrence matrix for your 100,000 types. Now you can associate each type with a point in a space of 100,000 dimensions – a rather breath-taking notion. 

That’s the general principle. Various methods can be used to reduce the number of dimensions we have to deal with, but regardless of how it is done we still end up with many more than three dimensions, which is the limit of what we can conveniently visualize. None of that matters to us. What matters to us is simply that we can associate the meanings of words with points in space and perform various operations in that space. 

Thus we have what has become the paradigmatic example for vector semantics [3]:

1) king - man + woman = queen

Other kinds of operation are possible:

2) paris - france + poland = warsaw

3) cars - car + apple = apples

All of these are based on analogies where a word is missing:

A : B :: C : ?

Remember, where you and I see a word the computer model sees a point in space, a point defined by a vector, which is a string of numbers. In each of these three cases it starts with a vector (that is, a string of numbers), subtracts a second vector from the first, then adds a third vector to that result, yielding a fourth vector, another point in space. The point is our answer. 

You and I may know the meanings of these words (examples 1 and 2) and the rules for pluralization (example 3), but the computer model knows none of that. It knows only the relations between word types that it can infer from the word-space constructed from the co-occurrence matrix based on the contexts in which word tokens occur. 

Gavin, however, isn’t interested in such analogies. Well, yeah, I rather suspect that he’s very interested in the fact that one can do such things with vector semantics, but that’s not what he does with the model he’s constructed. He uses it to examine a passage from Milton. 

As I mentioned at the outset, his co-occurrence matrix is based on 18,351 documents drawn from Early English Books Online. The structure in each of those documents must necessarily come from the mind of the document’s author. As those authors speak a common language we may, as I’ve argued above, think of the structure in that co-occurrence matrix as coming from the mind of a somewhat idealized corporate author of the corpus. Gavin is, in effect, asking that author to read a passage from Paradise Lost while we look on over their shoulder, as it were. 

That may not be how Gavin thinks of what he’s doing, but it’s how I’ve come to think of it. Call it a virtual reading.

Before the world of macro objects there was what? Quantum Chaos? [Tweet of the Day]

Yes, I agree with that. When an irreversible 'event' occurs and a wave that is spread out, it only shows up once in one place.

As a result there are no half dead/half alive cats --

The parable of Schrodinger's Cat ant the Law.https://t.co/aHpHCciN9q

— Traruh Synred (@synredae) December 3, 2018

So, the universe started as an immeasurable sea of quantum chaos and then somehow macro-objects emerged giving us time and space and change?

See Change seems more fundamental than time, August 21, 2018.

Powers of Ten [A Blast from the Bast]

If you've not seen this little gem, you should. It's marvelous. Note that that the most interesting and complex visual action centers of human-scale phenomena, a bit larger, at bit smaller, but human scale. At large and small scales the phenomena of the cosmos appear much alike. I've appended a note about that. (Originally posted in June 2011)

Some years ago I was rummaging through stacks in a used-book store in Troy, New York, when I came upon Powers of Ten: About the Relative Size of Things in the Universe. I smiled from ear to ear and bought it. I’d heard about he film for years, but never seen it, and I don’t recall whether I’d even known there was a book. But here it was in my hands, a book called Powers of Ten and it clearly was based on the movie.

What, you ask, is this marvelous movie/book about?

It’s about the whole universe. Let the Wikipedia explain:

Powers of Ten is a 1968 American documentary short film written and directed by Ray Eames and her husband, Charles Eames, rereleased in 1977. The film depicts the relative scale of the Universe in factors of ten (see also logarithmic scale and order of magnitude). The film is an adaptation of the 1957 book Cosmic View by Kees Boeke, and more recently is the basis of a new book version. Both adaptations, film and book, follow the form of the Boeke original, adding color and photography to the black and white drawings employed by Boeke in his seminal work.

So, two days ago I googled “powers of ten”, figuring that I’d come up with an online version of the movie somewhere. Well, I did better than that. The Eames Office has a website about the project. The website includes the film, but also photos and illustrations (including the originals from the book) and a blog. Lots of stuff there. Have fun.

* * * * *

This is a passage from William Benzon and David Hays, A Note on Why Natural Selection Leads to Complexity (Journal of Social and Biological Structures 13: 33-40, 1990):

Prigogine, who won the Nobel prize for demonstrating that order can arise by accident (Prigogine and Stengers 1984; Prigogine 1980; Nicolis and Prigogine 1977). He showed that when certain kinds of thermodynamic systems get far from equilibrium order can arise spontaneously. These systems include, but are not limited to, living systems. In general, so-called dissipative systems are such that small fluctuations can be amplified to the point where they change the behavior of the system. These systems have very large numbers of parts and the spontaneous order they exhibit arises on the macroscopic temporal and spatial scales of the whole system rather than on the microscopic temporal and spatial scales of its very many component parts. Further, since these processes are irreversible, it follows that time is not simply an empty vessel in which things just happen. The passage of time, rather, is intrinsic to physical process.

We live in a world in which “evolutionary processes leading to diversification and increasing complexity” are intrinsic to the inanimate as well as the animate world (Nicolis and Prigogine 1977: 1; see also Prigogine and Stengers 1984: 297-298). That this complexity is a complexity inherent in the fabric of the universe is indicated in a passage where Prigogine (1980: xv) asserts “that living systems are far-from-equilibrium objects separated by instabilities from the world of equilibrium and that living organisms are necessarily ‘large,’ macroscopic objects requiring a coherent state of matter in order to produce the complex biomolecules that make the perpetuation of life possible.” Here Prigogine asserts that organisms are macroscopic objects, implicitly contrasting them with microscopic objects.

Prigogine has noted that the twentieth century introduction of physical constants such as the speed of light and Planck's constant has given an absolute magnitude to physical events (Prigogine and Stengers 1984: 217-218). If the world were entirely Newtonian, then a velocity of 400,000 meters per second would be essentially the same as a velocity of 200,000 meters per second. That is not the universe in which we live. Similarly, a Newtonian atom would be a miniature solar system; but a real atom is quite different from a miniature solar system.

Physical scale makes a difference. The physical laws which apply at the atomic scale, and smaller, are not the same as those which apply to relatively large objects. That the pattern of physical law should change with scale, that is a complexity inherent in the fabric of the universe, that is a complexity which does not exist in a Newtonian universe. At the molecular level life is subject to the quantum mechanical laws of the micro-universe. But multi-celled organisms are large enough that, considered as homogeneous physical bodies, they exist in the macroscopic world of Newtonian mechanics. Life thus straddles a complexity which inheres in the very structure of the universe.

You can download the full article: Academia.edu, https://www.academia.edu/8488872/A_Note_on_Why_Natural_Selection_Leads_to_Complexity; SSRN, https://ssrn.com/abstract=1591788.

Tracks

Psycho-cultural distance as indicated by a newly calculated measure

This graph, for example, shows how the various populations score in relation to the United States and to China.

X-axis=distance from U.S. as measured by a battery of psychological measures
Y-axis=distance from China as measured by a battery of psychological measures pic.twitter.com/BJABHOchgs

— T. Greer (@Scholars_Stage) November 29, 2018

Based on work by Muthukrishna, Michael and Bell, Adrian and Henrich, Joseph and Curtin, Cameron and Gedranovich, Alexander and McInerney, Jason and Thue, Braden, Beyond WEIRD Psychology: Measuring and Mapping Scales of Cultural and Psychological Distance (October 2, 2018):

Abstract: We present a new tool that provides a means to measure the psychological and cultural distance between two societies and create a distance scale with any population as the point of comparison. Since psychological data is dominated by samples drawn from the United States or other WEIRD nations, this tool provides a “WEIRD scale” to assist researchers in systematically extending the existing database of psychological phenomena to more diverse and globally representative samples. As the extreme WEIRDness of the literature begins to dissolve, the tool will become more useful for designing, planning, and justifying a wide range of comparative psychological projects. We have made our code available and developed an online application for creating other scales (including the “Sino scale” also presented in this paper). We discuss regional diversity within nations showing the relative homogeneity of the United States. Finally, we use these scales to predict various psychological outcomes.

Available at SSRN: https://ssrn.com/abstract=3259613 or http://dx.doi.org/10.2139/ssrn.3259613.

Strange illumination

Vector semantics, you know: King - Man + Woman = Queen...

From MIT Technology Review, 2015:

The easiest way to think about words and how they can be added and subtracted like vectors is with an example. The most famous is the following: king – man + woman = queen. In other words, adding the vectors associated with the words king and woman while subtracting man is equal to the vector associated with queen. This describes a gender relationship.

Another example is: paris – france + poland = warsaw. In this case, the vector difference between paris and france captures the concept of capital city.

Baldwin and co ask how reliable this approach can be and how far it can be taken. To do this, they compare how vector relationships change according to the corpus of words studied. For example, do the same vector relationships work in the corpus of words from Wikipedia as in the corpus of words from Google News or Reuters English newswire?

To find out, they look at the vectors associated with a number of well-known relationships between classes of words. These include the relationship between an entity and its parts, for example airplane and cockpit; an action and the object it involves, such as hunt and deer; a noun and its collective noun such ant and army. They also include a range of grammatical links—a noun and its plural, such as dog and dogs, a verb and its past tense, such as know and knew; and a verb and its third person plural such as accept and accepts.

The results make for interesting reading. Baldwin and co say that the vectors sums captured in these relationships generally form tight clusters in the vector spaces associated with each corpus.

However, there are some interesting outliers where words have more than one meaning and so have ambiguous representations in these vectors spaces. Examples in the third person plural cluster include study and studies, run and runs, increase and increases, all words that can be nouns and verbs, which distorts their vectors in these spaces.

And it works in machine translation, but not always:

It’s worth noting that one of the pioneers and driving forces in this field is Google and its machine translation team. These guys have found that a vector relationship that appear in English generally also works in Spanish, German, Vietnamese, and indeed all languages.

That’s how Google does its machine translation. Essentially, it considers a sentence equivalent in two languages if its position in the vector spaces of each is the same. By this approach its traditional meaning is almost irrelevant.

But because of the idiosyncratic nature of language, there are numerous exceptions. It is these that cause the problems for machine translation algorithms.

See Ekaterina Vylomova, Laura Rimell, Trevor Cohn, Timothy Baldwin, Take and Took, Gaggle and Goose, Book and Read: Evaluating the Utility of Vector Differences for Lexical Relation Learning, arXiv:1509.01692v4 [cs.CL]:

Recent work on word embeddings has shown that simple vector subtraction over pre-trained embeddings is surprisingly effective at capturing different lexical relations, despite lacking explicit supervision. Prior work has evaluated this intriguing result using a word analogy prediction formulation and hand-selected relations, but the generality of the finding over a broader range of lexical relation types and different learning settings has not been evaluated. In this paper, we carry out such an evaluation in two learning settings: (1) spectral clustering to induce word relations, and (2) supervised learning to classify vector differences into relation types. We find that word embeddings capture a surprising amount of information, and that, under suitable supervised training, vector subtraction generalises well to a broad range of relations, including over unseen lexical items.

We came, we saw, we conquered. [Tweet of the day]

In six words or fewer, write a story about this photo... #amwriting #amreading #sixwordstory pic.twitter.com/GGikOoJLwR

— Lacey London (@thelaceylondon) November 30, 2018

In Latin: "Venimus, Vidimus, Vicimus."

Emotion among the Vikings as revealed in their literature

From The Medieval Review, Melissa Elmes reviews Rikhardsdottir, Sif. Emotion in Old Norse Literature: Translations, Voices, Contexts. Studies in Old Norse Literature. Woodbridge: D.S. Brewer, 2017. pp. 213. ISBN: 978-1-84384-470-9 (hardback).

The subject of emotion in premodern history and culture has enjoyed a rapid rise in critical interest and corresponding scholarship since at least the publication of Barbara Rosenwein's Anger's Past in 1998, although as Sif Rikhardsdottir points out in her introduction this critical attention has not to date extended enough to the subject of literary emotionality (14). [...]

In Old Norse/Icelandic studies there is, actually, a small but strong critical corpus on literary emotion; however, much of the scholarship to date on emotions in Old Norse/Icelandic (hereafter, ONI) texts has focused on the sagas and, within these studies, emphasis has been placed on the relative impassivity of saga characters--that is, the avoidance of emotion, rather than its narrative presence and function. Where emotions have been centered in such studies, they are centered in isolation and studied as individual categories--such as anger, laughter, and grief. [...]

The essential argument of this book is that the literary representation of emotion depends upon the development of an emotive script that is keyed to a particular set of societal expectations, and that that script in turn dictates through context (literary, historical, and/or cultural in nature) how the audience of a text interprets the emotions involved, both historically and in modern readership. Beginning with an Introduction that provides a brief overview of the history of emotions as a field of study, Rikhardsdottir moves into a discussion of the literary representation of emotion, identifying the traditionally-held view of fundamental difference between ONI sagas and romances as being grounded in a misunderstanding of how writers in these genres stage and express emotion. Rather, Rikhardsdottir locates in both genres a recognizable manipulation of emotional gestures for literary purpose--that is, a signposting of behavioral codes tied to literary convention, more so than actual social practice, that enables readers to infuse a given narrative space with emotional importance. This manipulation of emotional gestures constitutes emotive scripting which, in turn, readers must decode and activate in order to access the emotion in a given scene. Following this explanation of the concept of the emotional script, the rest of the book is organized into chapters spanning three categories of literature: the translated romances, the sagas, and the indigenous romances--that show precisely how the emotive scripts work in each genre, and across the ONI literary corpus overall.

Stagnation: Arnold Kling, "intangible factors are important determinants of economic outcomes."

 Arnold Kling comments on a recent conversation between Paul Krugman and Tyler Cowen. One of the topics that came up was increasing returns. Kling:

Context: This year, the Nobel Prize in economics went to Paul Romer and William Nordhaus, both of whom contributed to our understanding of economic growth. What facilitated the large increase in living standards in recent centuries, and what explains the large differences in living standards across countries?

Kling then gives us six short paragraphs of textbook background, worth reading. But I'm more interested in Kling's response:

In the video, Paul and Tyler point out some problems with the narrative of ideas and increasing returns. It seems to imply that economic growth will just get faster and faster, as we have more people combining more ideas. But measured economic growth, while still positive, appears to have slowed in recent decades. Also, since ideas can be used without being used up, why are some countries so backward in their use of ideas?

The answers to these sorts of questions take us out of the realm of typical economic factors. At one point, Paul quotes Robert Solow as saying that these discussions end up in a “blaze of amateur sociology.”

Solow’s name comes up in any discussion of economic growth. In the Solow model, the economic driver of productivity is savings. But there is a “residual” driver of economic growth, that Solow equates to pure advances in technology.

My own view is that intangible factors are important determinants of economic outcomes. I believe that they have become increasingly important in recent years. This limits our ability to explain economic growth on the basis of the measurable components of the Solow model.

Well, yeah, amateur sociology. But if that's where the answer is, then better look there than in professional economics, where it isn't. No?

And of course my favorite source of intangible factors is conceptual systems, specifically, the theory of cultural/cognitive ranks that David Hays and I worked on for a couple of decades in the previous century. There's nothing amateur about it, though it IS speculative. Anyone can ferret out facts using tried-and-true methods, even amateurs. It takes a real pro to speculate in a coherent manner. Here's what I said about speculation in the Preface to Beethoven's Anvil (p. xii):

Thus I like to think of this book as an exercise in speculative engineering. Engineering is about design and construction: How does the nervous system design and construct music? It is speculative because it must be. The purpose of speculation is to clarify thought. If the speculation itself is clear and well-founded, it will achieve its end even when it is wrong, and many of my speculations must surely be wrong. If I then ask you to consider them, not knowing how to separate the prescient speculations from the mistaken ones, it is because I am confident that we have the means to sort these matters out empirically. My aim is to produce ideas interesting, significant, and clear enough to justify the hard work of investigation, both through empirical studies and through computer simulation.

The importance of real-time processing of language

Erin S.Isbilen, Morten H. Christiansen, Nick Chater, It's about time: Adding processing to neuroemergentism, Journal of Neurolinguistics, Volume 49, February 2019, Pages 224-227, https://doi.org/10.1016/j.jneuroling.2018.04.005.

Linguistic exchanges occur in real time, on a moment-to-moment basis. The rapid rate of linguistic input (10-15 phonemes per second; Studdert-Kennedy, 1987), and its transience (50-100 ms; Elliott, 1962; Remez et al., 2010) pose a fundamental challenge to processing, with information being delivered at a rate that strains the limit of the human auditory threshold (∼10 non-speech sounds; Miller & Taylor, 1948). The additive effects of the linguistic signal's fast rate and fleeting nature are further exacerbated by the limitations of human working memory, which on average can retain no more than 41 (Cowan, 2001) to 7 [plus or minus] 2 items at a time (Miller, 1956). Together, these challenges form a Now-or-Never Bottleneck (Christiansen & Chater, 2016a,b): if input is not processed as soon as it is encountered, the signal is either overwritten or interfered with by new incoming material. In order to sustain linguistic functions, the cognitive system must overcome this bottleneck. Importantly, the Now-or-Never Bottleneck is not limited to linguistic processing. Rather, it extends to the perception of haptic (Gallace, Tan, & Spence, 2006), visual (Haber, 1983), and non-linguistic auditory input (Pavani & Turatto, 2008). Understanding how the cognitive system deals with this bottleneck can therefore provide fundamental insights into the emergence not only of language, but also of the other complex cognitive abilities discussed by HCRCSWY.

The dynamics of how the linguistic signal unfolds in real-time underscores the importance of memory processes in considering how the cognitive system deals with the Now-or-Never bottleneck. Building on the basic memory process of chunking, Christiansen and Chater (2016b) suggest that the cognitive system engages in Chunk-and-Pass Processing to overcome the bottleneck. Using Chunk-and-Pass Processing, the cognitive system builds a multi-level representation of incoming input, by rapidly compressing and recoding the input into chunks of increasing levels of abstraction as soon as it is encountered. This process of compression and abstraction enables information to be held in memory for longer periods of time. To provide an example from language, the raw acoustic input may be chunked into syllables, syllables into words or multi-word phrases, and so on up to complex representations of the discourse. Throughout the multi-level process of chunking, top-down information driven by predictions from semantic, pragmatic and discourse expectations augmented by real-world knowledge will enrich the resulting representations. The reverse is hypothesized to happen during language production, with the intended message being broken down into chunks of increasing specificity. [...]

From the viewpoint of the Chunk-and-Pass framework, language acquisition involves learning how to process input – that is, learning how to effectively chunk linguistic input using top-down information in the face of the Now-or-Never bottleneck. Importantly, the real-time pressures from language processing not only shapes language acquisition, but also the cultural evolution of language itself. [...]

Similarly, the incorporation of multiple cues in natural language can also facilitate both the usefulness and learnability of linguistic structures. Because the Now-or-Never Bottleneck makes back-tracking very hard, the language system needs to rely on all available information to be right-the-first-time when chunking the input.

And I'm guessing that it's the pressure of real-time processing that gives language its computational 'nature'.

Though we must  be careful here. I AM NOT asserting that computation is a basic or even the basic neural process. Rather, I am working within the scope of my conjecture that language processing is the most primitive form of computational process in the mind/brain. In particular, the mapping between linguistic form and meaningful content is where computation is necessary. What's computed, then, is the relation between chunks of form and chunks of meaning.

See my post, The Computational Envelope of Language, and the posts it cites as leading up to it.

Outside in the cold

That's from a graffiti battle that raged hot and heavy in Jersey City back in 2006 between Somer and a bunch of others. I tell a bit of that tale in this post, Tales Told on a Wall, A Year in the Life. There's more photographic evidence of the battle in this set: