True Magritte

Travelogue:  Not much to report here today as I’ve been focusing on work and haven’t gone out and about much yet this week.   I should have more to report after the weekend as I am going to see Moliere’s Les Femmes Savantes at Theatre du Nouveau Monde, en francais, and on Sunday my host has planned a tour of Montreal spots that I might not ordinarily get to as a tourist.   I might even drink some more beer.   In the meantime I have felt (strongly) my first earthquake (4.5, centered just southeast of downtown) and shared an enjoyable lunch with an old college friend who is an economics professor at the University of Ottawa.  Last night I attended a talk at Concordia given by Dr. William McComas of, holy mother of poutine, the University of Arkansas.   I refrained from saluting him with a Hog call.    

But here are a couple of pictures of where I’ve been spending most of my time.

The view from my apartment on Boulevard de Maisonneuve

Looking down Ste. Catherine 

Pictures of the Concordia-Loyola campus where I am working to follow.

Psychologue:  Ever since I saw an exhibit at SFMoMA in 2000, I have been fascinated with the art of Rene Magritte, the Belgian surrealist/absurdist painter.     If no others, you are probably familiar with these works of his entitled, “Le Fils de l’Homme”  (The Son of Man) and “la trahison des images “ (The Treachery of Images).

As I discovered at that exhibition for the first time, Magritte painted a number of additional works that build on the theme of the famous “Ceci n’est pas une pipe.”   In these works, images are paired with  written words that apparently serves to label the object depicted in the images.   Here are a couple.  

     So Magritte either had severe anomia (none of the labels match the images) or was onto something really important.   I’m going with the latter, but whereas Magritte was mostly concerned with treacherous images, I view these works as an excellent illustration of the arbitrary relationship between words and the objects they label.   As I said in a previous entry, there is no clear causal connection between the phonological shape of a word and that word’s referent.   So what’s treacherous here—the image, the word, or the association between the two?

     This work is also relevant to research on early word learning that has been important in my field in recent years.  This work on ‘selective trust,’ pioneered by my colleagues Melissa Koenig and Paul Harris, has shown that children as young as 3 prefer to learn from somebody who has a history of being right about things.   So, in a typical selective trust experiment, children are exposed to two speakers, one who labels common objects correctly and one who labels them incorrectly.   A novel object is then presented and children are either asked to choose which of the speakers they would ask to find out what the object is called or, more typically, they hear the speakers provide different novel labels for the new objects and then are asked what it is called (“What do you think?  Is this a modi or a gimble?”).   Some three-year-olds and most four-year-olds choose (the word used by) the speaker who’s been reliable in the past to guide these judgments.   There are many diverse demonstrations of this effect, and my former student Jason Scofield and I have even shown that 4-year-olds can do this retroactively; i.e. they learn a word and then find out later that the speaker doesn’t know a shoe from a pencil and are then given the opportunity to choose the label they originally learned or a label provided by another speaker.   Most 4-year-olds switch to the other label.      

     That’s all well and good, but how about this Magritte gem (ignore the colorful border which is not part of the original):

     Rene Magritte was a native French speaker but here he is labeling (incorrectly) common objects in English.   This is an interesting case because perhaps we shouldn’t expect Magritte or any francophone to know the correct English labels for these objects.  So this raises the following question:  Whom would you rather learn from, a speaker of your own language who doesn’t know a shoe from a pencil or a speaker of another language that you may not understand but who may know what things are called?  This is a question I’ve asked in another series of studies with some of my students.  Using the selective trust procedure, we initially found that preschoolers (in the US) prefer to learn from an inaccurate English speaker to an accurate Spanish speaker.   Of course, these kids didn’t understand Spanish so they didn’t know whether the speaker was accurate, inaccurate, or speaking gibberish so in a second study we provided a translation of the Spanish word  (e.g. “zapato is the Spanish word for shoe”) and the children changed their preference for word learning from the inaccurate English speaker to the accurate Spanish speaker.  

     This is a cool finding because we  find that in nearly all social situations young children prefer same-language speakers (even preferring unaccented speech to accented speech as shown by Kinzler and her colleagues) but here, for knowledge-based tasks, accuracy trumps language.   A follow-up question that could be asked here in Montreal is how bilingual children perform on these tasks.   Finally, an interesting question is whether social preferences would change as well in this situation.   Think about it, would you rather hang out with someone who speaks a different language but knows his stuff, or with someone who speaks your language who doesn’t know his trou from his Magritte?  I'd like to know your gut reaction to this question. We have some data on this question from children, so stay tuned.