September 17, 2007

Survived by his trainer

In case you missed them, there were a couple of updates added to my report of the death of Alex the African Grey parrot. I also have a couple more comments of my own, so I'm consolidating all these into this post.

First, the updates: a Language Log reader wrote to inform us about an interesting discrepancy between Alex's final words as reported in the NYT story that I cited and this Boston Globe story:

NYT: Even up through last week, Alex was working with Dr. Pepperberg on compound words and hard-to-pronounce words. As she put him into his cage for the night last Thursday, Alex looked at her and said: "You be good, see you tomorrow. I love you."

Boston Globe: Pepperberg said she and Alex went through their good-night routine, in which she told him it was time to go in the cage and said: "You be good. I love you. Iíll see you tomorrow." To which Alex said, "You'll be in tomorrow."

I think it would be no more or less impressive for Alex to have predicted Dr. Pepperberg's return in the morning than for him to have expressed his loving feelings for her. But would the prediction-of-return have generated anything like the following editorial musings on the expression-of-love? (Pointed out by the same LL reader.)

These are bottomless questions, of course. For us, language is everything because we know ourselves in it. Alex's final words were: "I love you."

There is no doubt that Alex had a keen awareness of the situations in which that sentence is appropriate -- that is, at the end of a message at the end of the day. But to say whether Alex loved the human who taught him, we'd have to know if he had a separate conceptual grasp of what love is, which is different from understanding the context in which the word occurs. By any performative standard -- knowing how to use the word properly -- Alex loved Dr. Pepperberg.

Now my own updates. The NYT story I originally cited was "Alex, a Parrot Who Had a Way With Words, Dies" (by Benedict Carey, published Sept. 10). As often happens, this story was lightly edited and republished the next day with a new title significantly referencing the expression-of-love: "Brainy Parrot Dies, Emotive to the End". Among the edits are these:

  1. "demonstrated off", which I figured must have been an editing error in my original post, has been changed to "demonstrated". (In the same sentence, "programs on the BBC and PBS" has been changed to "programs on PBS and the BBC" for some reason -- America first?)
  2. A sentence originally ending "... scientists had little expectation that any bird could learn to communicate with humans" now ends "... scientists had little expectation that any bird could learn to communicate with humans, as opposed to just mimicking words and sounds".

In yesterday's Week in Review, George Johnson has a piece provocatively titled "Alex Wanted a Cracker, but Did He Want One?" There you can see a YouTube video showing off (or demonstrating off, if you prefer) some of Alex's purported skills with language and numbers. (Johnson also correctly identifies the PBS show hosted by Alan Alda that Alex appeared on: it's "Scientific American Frontiers", not "Look Who's Talking" as Carey had reported.)

As of this writing, the NYT sidebar indicates that Johnson's is the 8th most e-mailed NYT piece -- but note that the italics are missing from the title, unfortunately disguising what I find to be the most interesting aspect of the piece.

This most interesting aspect is the nature of consciousness, and whether Dr. Pepperberg's research with Alex (and other African Grey parrots) has anything to tell us about it. This is touched on in the two paragraphs that sandwich a longer passage about Alex's purported understanding of the complex mathematical concept of "zero":

[Dr. Pepperberg] is quick to concede the impossibility of proving that the bird was actually verbalizing its internal deliberations. Only Alex knew for sure.

[...]

In a well-known essay, "What Is it Like to Be a Bat?" the philosopher Thomas Nagel speculated about the elusiveness of subjectivity. What was it like to be Alex that last night in his cage? We'll never know whether there really was a mind in there -- slogging its way from the absence of a cork-nut to the absence of Alex, grasping at the zeroness of death.

Nagel's article is well-worth reading. The original is available from JSTOR, but for those without JSTOR access there several alternatives -- see this or this, the short summary and discussion here, or if you prefer PDF to HTML, try this. I happen to have read the article myself this summer as reprinted in Hofstadter & Dennett's The Mind's I: Fantasies and Reflections on Self & Soul (which I also highly recommend to anyone interested in the nature of consciousness).

A final comment. At Dr. Pepperberg's Alex Foundation website, the news page has the following announcement:

Your kind attention please.

We have had a many orders placed recently and it may be a few days to get everything back up to speed. Some of the shipping costs are incorrect we were working on the issue, before the events of Sept 7th.

What struck me here was not the mistake in the last sentence ("...shipping costs are incorrect we were working..."), no doubt brought about by a foundation in chaos after the death of its primary subject. (I've been taken to task for pointing out these kinds of mistakes before.) I'm more interested in the very last noun phrase: "the events of Sept 7th". This is a reference to Alex's death and the ensuing media attention, of course: Alex was found dead on Friday the 7th (though note that the foundation's official press release notes the date of death as Thursday the 6th). When I read or hear the phrasing "the events of [month date]", I invariably think of the terrorist attacks of September 11, 2001 -- and so of course it doesn't help here that the month is September.

In a way, Google agrees with me. Searching for the strings {"the events of month"} returns significantly many, many more hits when month = sept(ember) than any other month -- in fact, about 3.5 times all other months put together. And not surprisingly, paging through the sept(ember) results reveals almost nothing but references to the terrorist attacks.

the events of ... jan(uary) feb(ruary) mar(ch) apr(il) may jun(e)
ghits 39,500 31,930 45,070 57,604 85,400 51,108
the events of ... jul(y) aug(ust) sept(ember) oct(ober) nov(ember) dec(ember)
ghits 56,302 55,130 1,985,000 55,070 42,202 46,970

Interestingly, each of the other months are in a rather tight range, with the mean and the median both around 51K ... interesting, but I've already invested too much time today (indeed, this month) writing about a dead parrot.

[ Comments? ]

Posted by Eric Bakovic at September 17, 2007 03:27 PM