Tuesday, 24 June 2014

Shake Can Well (Link)

Last month, I gave a talk at CamCos3, the text of which is now posted on LingBuzz. It was the first time I have ever given a theoretical talk in Cambridge, let alone an invited one. And by happy coincidence, this year marks the 30th anniversary of my graduation from Trinity Hall. All in all, a very pleasant and interesting homecoming.


Monday, 27 January 2014

Names and Faces, and Handedness (Preliminary Results)

Who would've guessed...? After years of not getting round to it, then (literally) months of preparation, I've been able over the past few weeks to run a new Language and Cognition experiment. Here are some preliminary results...

The study I've set up investigates possible correlations in "perceptual narrowing", comparing participants' ability to discriminate native vs. non-native phonemic contrasts with their ability to discriminate native vs. non-native Japanese vs. Caucasian faces (for anyone interested, there's some background information on my website). The experiment consists of two forced-choice discrimination tasks—XAB/ABX, for faces and sounds, respectively—plus a lexical familiarity pretest (to test whether participants are familiar with all of the English words carrying the phonemic contrasts).


To date, 43 Japanese undergraduate students have taken part in the first run of the experiment. The results below are based on 40 participants (including 4 left-handers: see below). Two participants, whose accuracy scores fell below two standard deviations from the overall mean, were dropped from the analysis. One other participant (subject #3), who reported spending her early childhood in the US, was also excluded from the present analysis.


Apriori, one might have expected that (for Japanese participants) Japanese faces would be easier to discriminate than Caucasian faces. Right? Except, that's not what the data suggest (at first blush). See Figures 1 and 2. As these figures show, the Japanese faces elicit somewhat less accurate and slower responses, though it remains to be seen whether the differences are statistically reliable. (Of course, this might not be a fatal problem: it could just be that the 30 pairs of Western faces are just easier for contingent reasons. It depends on how the Western participants judge the two sets, and I won't find this out until April. In any case,  I'm mainly interested in probing cross-modal correlations within-subjects, so even if the overall results don't turn out quite as expected, there may still be interesting within subject correlations.

Figure 1. Overall, Asian faces are hardest to discriminate (though the effect may not be robust)...
Figure 2...and Japanese participants take longer to judge Japanese faces than Caucasian ones (overall).

At this point in the analysis, the interesting quirks come when one looks more closely. Among the other independent and control variables, besides NATIVENESS, I manipulated ISI—time between the offset of the target face and presentation of the two possible matches (500 vs. 1000ms), Presentation Time of the target face (100 vs. 200) msecs, and Orientation—whether the faces are presented upright or inverted (upside-down). I also recorded the between-subjects factor of Handedness, just because I have a hunch about this...

Taking Orientation first, Figure 3 shows that this matters a lot: as predicted, in every condition, upright faces are easier to identify than the same faces presented upside-down.  No surprise here. What is surprising, though, is that the inversion effect is probably only significant for Asian faces (in other words, for native faces!). Curiously, inverted Caucasian faces are not much harder to discriminate than when presented upright.

Figure 3. Interactions between Nativeness and Inversion. The chart here shows that—as expected—upright faces are easier to discriminate than inverted ones. No surprise here. What is surprising, though, is that the inversion effect is probably only significant for Asian faces (in other words, for native faces)

I haven't even started the interpretive stats yet, so some of these initial results may turn out to be spurious, but it looks like all of these factors interact in interesting ways. Perhaps most surprisingly, handedness seems to interact with the other variables in an interesting way: left-handers are better than right-handers overall—much better in certain conditions—but also appreciably worse in others. See Figures 4 and 5 below. (Less surprising was the apparently significant main effect of Orientation, though it is still curious that this effect is larger for left-handers than right-handers.)

Figure 4. As might have been predicted, it's harder—for everyone—to discriminate faces overall when they are upside down (INV) than the right way up. That's what we see in Figure 3. What is less expected are the sub-patterns revealed in Figure 4: left-handers seem to show a huge effect of orientation for native (Japanese) faces, but not for non-native (Caucasian) ones; right-handers, by contrast, are better overall at discriminating non-native faces, but relatively poor at discriminating native ones.
Figure 5. Shorter ISI results in greater accuracy, but only for native faces (for left-handers) and for non-native faces for (right-handers). Go figure!

Which is all very weird, until I looked again and realized that there were only 4 left-handed subjects in the set (42). So it may really be a freak result. More anon.


As for the sounds, these have worked out almost exactly as predicted. Figures 6 and 7 shows the accuracy scores and response latencies to trials involving minimal phonemic contrasts in English and Japanese, embedded in real lexical items. The blue bars show the discriminability of the control items: /t/ vs. /d/, phonemic contrasts common to both languages. The red bars show participants' performance on test items, i.e. contrasts that are not shared by both languages: /l/ vs. /r/ in the English trials, phonemically short vs. long segments in the case of the Japanese trials. Within the test items, 'more difficult' English contrasts are those where the critical phoneme is non-initial in the syllable (e.g. prays/plays), in contrast to the more perceptible syllable-initial contrasts (low/row); for the Japanese items, differences in vowel length were predicted to be harder to detect than consonantal length contrasts.

Figures 6 and 7 clearly suggest main effects of LANGUAGE (English vs. Japanese) and TRIAL TYPE (Control vs. Test), also—possibly—a small main effect of ITEM DIFFICULTY, as well as a reliable interaction between language and trial type, as reflected in both dependent measures: the Japanese participants in the study are significantly less accurate and slower in correctly discriminating non-native phonemic contrasts—i.e., l vs. r—than in discriminating native ones (67% vs. 98%). Notice that there is no appreciable difference in the responses to English and Japanese control items (both >95%).

Figure 6. Sounds—Accuracy across Conditions
Figure 7. Sounds—RT across conditions

The sounds experiment also manipulated ISI, in this case, the delay between presentation of the second alternative sound and the target sound (either 2250 or 3250 msecs). Ceteris paribus, longer ISIs should increase task difficulty. However, as Figures 8 and 9 show, increasing ISI appears to have had no significant effect on overall accuracy, though there was a (possibly significant) increase in response latency in the English/Test Conditions. 

Figure 9. Sounds. Effects of ISI on response latency (Note that the RTs here are based only on correct responses.)

Sunday, 17 November 2013

Kanji, how I loathe thee!

Events of the last few weeks have led me to give some consideration to Japanese Kanji, and its moraic sidekicks, Katakana and Hiragana. For what it's worth: here's my dystopic theory of the Japanese writing systems.

The purpose of Kanji is: 
  • (i) to prevent (almost) anyone who has not been through six years of elementary education from learning the language, by blocking all feedback from written signs and messages; 
  • (ii) to waste children's time by focusing their attention on the brushstrokes of arbitrary symbols , when they could be learning content; 
  • (iii) to cause untold domestic and other friction between those who have already learned a few kanji, and those who are just starting out...

The contrasting purpose of katakana is: 

  • (i) to stymie the best efforts of Japanese speakers trying to learn English, by neutralizing crucial distinctions (not just light/right but also lunch/ranch), inserting superfluous syllables, and generating atrocious puns; 
  • (ii) to offend against the (a)esthetic sensibilities of almost everyone (except, perhaps, those who liked thin, angular cars like the Triumph TR7, or late-70s Ford Escort).
And then there's hiragana, whose purpose is much less clear, unless it be:
  • (i) to teach Japanese linguistics students about functional categories; 
  • (ii) indicate at a distance where one can buy cigarettes (by a curious quirk of history *tobacco* is invariably written in hiragana, rather than the expected katakana.) 
These would seem to be rather petty uses for a perfectly serviceable, almost attractive syllabary.
End of rant. 

Which is not to pretend that English orthography is perfect. Vietnamese is nice, though.

Wednesday, 23 October 2013

Der Schauende (cross-posting)

This will not be news to those of you who had a decent mathematical education, but it was to me. After 51 years, thanks to my son Sean's science homework, I've just learned that there is a difference between precision and accuracy. Really, I had no idea...

(Accuracy, it seems, is only for realists: but even relativists can be—relatively—precise.)

Now for some catch-up.

(Same series, though mine is an earlier edition)
A few months ago, I picked off the shelf a copy of Rilke's 'Der augewählten Gedichte, anderer Teil. (I don't have the erster Teil, unfortunately). This undated Insel Verlag edition was published in Leipzig, either just before or during the War, judging from the cover, and must have belonged to my uncle Kenneth (erstwhile Professor of German at Keele University), as do all of the few decent German books I now possess—including an early copy of Hermann Paul's Prinzipien der Sprachgeschichte, the translation of which will be on my to-do list, probably for ever. It won't happen today, that's for sure: I taught four 90-minute classes virtually back to back, without lunch, came home and cooked dinner, helped to bathe and put the kids in bed...

...Anyway—und dann und wann ein weißer Elephant' (which isn't in this collection, but still careers around my head at least once a week)—I was at my reading of 'Der Schauende' some months ago, and thought: Wow (or some such hint at the ineffable).

What a brilliant piece.

...Wie ist das klein, womit wir ringen
Was mit uns ringt, wie ist das groß:
ließen wir, ähnlicher den Dingen,
uns so vom großen Sturm bezwingen — 
wir würden weit und namenlos.

Was wir besiegen, ist das Kleine
und der Erfolg selbst macht uns klein...

Stunning, really. Disheartening, too: after hearing these too perfect lines, writing anything—saying anything—seems only to add more junk to an already overflowing landfill.

And so I looked on YouTube for a competent rendering of the poem, and found something quite dreadful, an object lesson in how not to read anything (not even a grocery list), in how a bad performance can utterly ruin even the finest lines.

But then I looked again today
And found the poem resurrect
by Oskar Werner, a man whose name,
though t'is renowned, I never knew
Till now — and more's the shame.

[with apologies to Heine]

YouTube taketh away, but giveth, too
...und dann und wann: ein Wunder — Click to play

Was wir besiegen, ist das Kleine
und der Erfolg selbst macht uns klein...

Postscript 10/25: This evening, browsing my subscriptions on Youtube, I came across this talk by Nigel Fabb (one of the other Nigel linguists) on verse design and verse delivery, focussing on Dylan Thomas' reading of his own and other's poetry. Fascinating material, interesting ideas:

Tuesday, 11 June 2013

Linguistics of Vietnamese

After some delay, I am pleased to have this in my hands at last. Thanks to Daniel Hole and Elisabeth Löbel, and the staff at Mouton for doing such an excellent job. If you are interested in Vietnamese grammar, please have your library order this!

A preprint version of the article is available here:

Wednesday, 17 April 2013

One Direction: Down

As mentioned in another post, one of the classes I teach at Konan is called Kiso-enshu I, which roughly translates as Introductory Seminar, or so I'm told. There is no prescribed curriculum or syllabus for this course: it is intended an introduction to academic studies for incoming students of English, who until just a few weeks ago were still in high school.

The gap year is still alien to most Japanese students, more's the pity.

In the absence of a fixed curriculum, I am free to do "pretty much what [I] like". Since the intersection of [the set of activities covered by this expression] and [the set of legitimate and appropriate things to do with a class of slightly post-adolescent teenagers] includes listening to and talking about English songs—and not much else—that is what we are doing.

Beginning next week, the course will introduce students to singer-songwriters that, but for this course, they would never, ever, listen to: Harry Chapin, Ralph McTell, Joni Mitchell, Don McLean, and Leonard Cohen are up there, for starters. They may not like it, but I am a man on a mission. And missionaries—at least in the popular Victorian stereotype—need to find out and understand what kinds of hideous alien gods their charges are currently in thrall to. It's important to have a base-line reference. So last week I asked my students to fill in a questionnaire about their favourite English-speaking bands.

A small, but marginally significant group among those who expressed a preference wrote 1D (= One Direction). I had heard of this group, of course—my 5 year-old nephew is quite a fan, but until yesterday, when I started to prepare for today's class, I had no idea—really n o  i d e a—how bad, how jaw-droppingly, bletheringly, numbingly awful it could be. It is sometimes said of really bad art that "it's so bad, it's great", but this doesn't apply to 1D: the progression from awful through excruciating to hysterically revolting is completely linear (as their name suggests). What is most striking in the one song I assigned myself and transcribed below is not simply the complete absence of any musical or lyrical talent, considered separately: it is the almost surreal lack of correspondence between natural English prosody and musical metre that is so wretched-making. If the proverbial monkey were given a pen and asked to put stress on random syllables in each line, s/he could not have produced more unnatural-sounding English or more forced metre. This is not mere doggerel, it is much, much worse than that: indeed, for the writer of this song, doggerel must be an aspirational goal, rather than a pitfall to be avoided.

As evidence I offer the following specimens. Specimen 1 and 2 present two transcripts of the song What makes you beautiful: Specimen 1 contains only the syllables in each line that receive strong stress as sung by 1D; Specimen 2 contains the syllables that should receive stress if the song were read as a rhyming text. For purposes of confirmation only, a link to the VEVO video is embedded below.

Be warned that this may be injurious to your health: it is certainly not pretty.

Specimen 1.

heads walk do-o-r
(make-)up, up, way are en-ou-ou-gh
Every else room, you
light world body
you hair over
smile ground, hard
don't kno-o-ow, don't know you're beautiful
If you I, understand want  des(perately),
looking can't believe
kn-o-o-ow, You don't know you're beautiful,  
on, wrong,
right, it, song
why, shy, away, look, eye eye eyes...

Specimen 2. Most egregious forms highlighted

You're insecure, Don't know what for,
You're turning heads when you walk through the door,
Don't need make-up, To cover up
Being the way that you are is enough

Everyone else in the room can see it, Everyone else but you:
Baby, you light up my world like no-body else
The way that you flip your hair gets me overwhelmed,
But when you smile at the ground it ain't hard to tell, You don't know, Oh, oh,

You don't know you're beautiful,
If only you saw what I can see,
You'd understand why I want you so desperately,
Right now I'm looking at you and I can't believe,
You don't know,
Oh, oh, You don't know you're beautiful,
Oh, oh, That's what makes you beautiful.

So come on,
You got it wrong,
To prove I'm right, I put it in a song,
I don't know why you're being shy,
And turn away when I look into your eyes

This is not a question of vocabulary choice. As Something shows, one can make a song out of simple words that still manages to read almost exactly as well as it is sung. All you need is...love talent, something that these boyos and their aesthetically bereft team just haven't got.

Jesus wept.

Something in the way she moves
Attracts me like no other lover
Something in the way she woos me
I don't want to leave her now
You know I believe and how

Somewhere in her smile she knows
That I don't no other lover
Something in her style that shows me
I don't want to leave her now
You know I believe and how

You're asking me will my love grow
I don't know, I don't know
You stick around now, it may show
I don't know, I don't know

Something in the way she knows
And all I have to do is think of her
Something in the things she shows me
I don't want to leave her now
You know I believe and how

Friday, 5 April 2013

Minimalism and Semantic Syntax: Interpreting Multifunctionality in Vietnamese

Full Paper Here

Talk presented in Hanoi, May 11-12, 2013

Original Abstract
This talk is concerned with a deceptively simple question: where does sentence meaning come from? Within generative grammar, at least since the demise of Generative Semantics, the received view has been that the meaning of a sentence is exhaustively a function of the lexical elements of which it is comprised (setting aside the effects of constituency and scope). This is made explicit in the Projection Principle (Chomsky 1981), which understands syntax as a "projection of lexical properties". In subsequent Minimalist approaches (Chomsky 1993, 1995, 2000), this restriction is tightened up even further by the requirement that syntactic computations operate exclusively with the lexical items introduced in the initial array (numeration): no node labels or extraneous symbols (e.g., theta-roles, indices, movement traces, levels of representation) which might contribute to sentence meaning. This does not, of course, exclude reference to abstract formal features—indeed, these are crucial to most Minimalist analyses—but it requires that such features (e.g., EPP features) are ultimately drawn from the lexicon: they are themselves lexical entries, alongside contentful, arbitrary lexical items. Whatever the theoretical advantages of this approach for delivering an extremely spare Minimalist syntax, it should be clear that it massively increases lexical complexity, leading to a multiplicity of different abstract features attaching to what are, intuitively speaking, the same lexical items. Grammatical theory is a ‘zero-sum game’: if the syntax does little or no semantic work, the burden necessarily falls on lexical specification.

In the case of languages with rich inflectional paradigms and/or an extensive inventory of phonetically-differentiated functional categories, this 'poor syntax—rich lexicon' approach makes some sense, since subtle differences in feature specification are reflected in different pronunciations that must in any case be lexically listed; e.g., English present perfect has been vs. preterite was; wh-interrogative who vs. indefinite anyone; locative vs. expletive there; nominative she vs. accusative her. However, for Vietnamese and other isolating languages, the desirability of a strict lexicalist approach is much less evident. In contrast to inflectional languages, Vietnamese does not appear to differentiate subtle meaning contrasts in the lexicon: instead, it disposes of a set of radically-underspecified 'multifunctional' items, whose semantics are determined in part—and in some instances exhaustively—by their position in phrase-structure. A clear example of this multifunctionality is offered by the modal auxiliary được (also phải), which is variously interpreted as a deontic, epistemic or abilitative modal—even as a non-modal, aspectual, particle—in different structural positions, This is illustrated in (0); see Duffield (1999), Phan & Duffield (in prep.)

0. a. Ông Quang được mua cái nhà.
        prn Q. can buy cl house
        ‘Quang was allowed to buy a house.’

    b. Ông Quang mua được cái nhà.
        prn Q. buy can cl house
        ‘Quang has bought (was able to buy) a house.’

    c. Ông Quang mua cái nhà được.
         prn Q. buy cls house can
         ‘Quang is able to buy a house/Quang may possibly buy a house.’

Other examples will be discussed directly. This multifunctionality suggests a radically different, though equally austere, conception of Minimalism: Minimalist Lexicalism (see also Marantz 2005, Borer 2007). The corollary of this, of course, is Semantic Syntax: meaning inheres in, and is read off of, syntactic representations. In this talk, then, I elaborate an alternative Minimalist thesis: I argue that it is elucidating to introduce a limited amount of meaning into syntax, maintaining that this can be done without resurrecting Generative Semantics.