The Tristesse of No Bonjour

July 27, 2018
Posted by Jay Livingston

Several years ago, I went to Paris with my family. When we got out of the airport, I couldn’t find the RER, the express train from DeGaulle to Paris. I went up to a man standing on the sidewalk and asked, in French of course.

 “Bonjour,” he said.

I repeated my question. “Bonjour,” he said again, this time as if cuing a dim-witted child. I got it. “Bonjour,” I said and again asked about the RER. This time he answered.

I had chalked it up to this guy just being a stickler for formalities. But now that I’ve started reading The Bonjour Effect by Julie Barlow and Jean-Benoît Nadeau, I realize how wrong I was. Bonjour is not just a greeting. It’s like eye contact – a necessary start to any interaction. It acknowledges that you are in the same situation with the other person. Without bonjour, communication cannot begin.

As I read this first chapter about bonjour, I recalled a much earlier visit to Paris. I needed some Velcro to make a small repair on something. A piece of clothing? A bag? I don’t recall. A friend told me that I could find Velcro in the mercerie section of a department store.  I went to La Samaritaine and found the mercerie. Two sales girls were standing talking to each other. I stood there, waiting to be waited on. Any clerk in an American store would have turned to me and asked if she could help. But the two girls continued their conversation, facing one another and ignoring me as if I weren’t there. I can’t remember how I managed to interrupt and finally get the Velcro.*

The rudeness of the French, I thought, or at least young French women. But now, decades later, I wonder what would have happened if I had said, “Bonjour.”

Of course, it’s not just a matter of words. The bonjour requirement is the visible tip of an underlying difference in the way we think about service workers and customers and the relation between them. The definition of those roles in France is not the same as it is in America. Barlow and Nadeau explain:

When you enter a French store ore a restaurant or even walk up to an information kiosk, the first thing you have to do in France is acknowledge that you are entering their turf. That’s because you are asking for something from an employee who may have something more important to do. Whether or not that employee actually does have something better to do is not the point. You are interrupting him to ask for something. He does not owe you anything in exchange for you giving him your bounces. The French just don’t think that way. When you address a merchant or a clerk or a hostess or even a waiter, bonjour is not a word. It’s not a greeting or even a form of courtesy. Bonjour is code for “please allow me to indulge in your services.”

---------------------
* The French word for Velcro, I discovered, is Velcro. It was invented by a francophone Swiss. According to Wikipedia, the word is a portmanteau of velour and crochet (hook).

Am I Blue? It Depends.

July 25, 2018
Posted by Jay Livingston

In olden days a glimpse of stocking
Was looked on as something shocking
But now, God knows,
Anything goes.
   — Cole Porter

“Defining Deviancy Down” is the title of a 1993 article by Daniel Patrick Moynihan. Like Cole Porter, Moynihan was arguing that behavior looked on as shocking in earlier times nowadays gets a pass. But while Cole Porter was bemused, Moynihan was deeply worried.  The subtitle of the article was “How We've Become Accustomed to Alarming Levels Of Crime and Destructive Behavior.”

Moynihan begins with two related ideas from sociology. First, norms change. Second, whether a behavior is deviant depends on how much of it there is. Society seems to need a certain amount of deviance, and when behavior changes, the norms change so as to maintain that amount. Moynihan quotes from Wayward Puritans, Erik Erikson’s classic study of deviance in colonial New England. “the number of deviant offenders a community can afford to recognize is likely to remain stable over time.”

Erikson in turn was inspired by Durkheim’s well-known quote about the necessity of deviance.*

Imagine a society of saints, a perfect cloister of exemplary individuals. Crimes, properly so called, will there be unknown, but faults which appear venial to the layman will create there the same scandal that the ordinary offense does in ordinary consciousness. If, then, this society has the power to judge and punish, it will define these acts as criminal and will treat them as such. For the same reason, the perfect and upright man judges his smaller failings with a severity that the majority reserve for acts more truly in the nature of an offense. Formerly, acts of violence against persons were more frequent than they are today, because respect for individual dignity was less strong. As this has increased, these crimes have become more rare; and also, many acts violating this sentiment have been introduced into the penal law which were not included there in primitive times.

Both Durkheim and Erikson focused on societies that defined deviance up. Durkheim’s society of saints was hypothetical, but Erikson’s was real. Those 17th -century Puritans in Salem thought of themselves, or at least their leaders, as saints. Three hundred years later in the US, a dearth of sinners was no longer the problem, as Moynihan saw it. Just the opposite. But the general proposition is the same: the amount of a behavior affects our perception of how deviant it is.

As with deviance, so with color. How blue does a dot have to be for you to say that it’s blue and not purple? The answer, according to a recent series of experiments, is that it depends. It depends on the actual color of the dot, of course. That’s what the Graph A shows. The researchers (Levari, Gilbert, Wilson, Sievers, Amodio and Wheatley) asked subjects whether a dot was blue or purple. Subjects (I hope they were well paid) had to judge 400 dots. In the graphs below, the X-axis is the degree of blueness.

(Click on the image for a larger view.)

The bluer the dot, the higher the percentage of subjects who labeled it as blue. In one version of the experiment (Graph A) the first 200 trials and the last 200 had the same proportion of blue dots. Consequently, color was everything, First 200 trials (blue line) or last 200 trials (yellow line) made no difference.
                                                                               
But in the other version of the experiment, in the final 200 trials, the experimenters reduced the number of really blue dots. As in the first experiment, color made a big difference. Bluer dots got a higher percent of subjects calling them blue. But the number of blue dots also affected perceptions. There was little effect for the dots that were at the ends of the spectrum — the very purple and the very blue.

But for the more ambiguous colors, the number of blue dots affect subject’s willingness to call a dot blue. I’ve added red boxes to show more clearly the difference for a single level of blueness. In the first 200 trials, less than half the subjects called those dots blue. But with fewer of those dots around, subjects were far more likely (more than 75%) to define that color as blue.

There’s something else interesting in Graph B — those five yellow data points indicating that when blue dots were scarce (the last 200 trials), even the very blue dots were not always labeled as blue. In some cases, only half the subjects saw them as blue. Apparently, when blueness (or deviance?) becomes rarer, there is less consensus on just what is and isn’t blue (or deviant).

Some of the experiments involved questions far more subjective than “Is this dot blue?” for example, “Is this research proposal ethical?” The results were similar. This tendency to define deviance up, the authors say, leads to an irony that the people involved often do not see: “well-meaning agents may sometimes fail to recognize the success of their own efforts.” Take “micro-aggressions” on campus, for example. You can see these as an important problem, one that requires constant vigilance and action. But you can also see their elevation to the status of “problem” as a sign that the more egregious bright-blue-dot forms of sexism and racism have grown scarcer.


------------------------------
* Somehow, the writers of the TV show “Profiler” managed to have Debbie, their sociology grad student character, completely miss Durkheim’s basic idea. See this earlier post.

Camille — a Name That’s Bucking the Trend (in France)

July 19, 2018
Posted by Jay Livingston

Harper, Avery, Aubrey, Riley, Addison were among the most popular fifty names for girls last year. These fit a general pattern — first they are names for boys, then become acceptable and often stylish for girls.

Often, once a name has crossed the gender line, parents of boys find it less and less attractive. In an earlier post (here), I referred to this as the “there goes the neighborhood” effect. The lower-status group (in this case girls) move in, the higher-status group leaves. And they don’t come back.

Here’s Aubrey:

 

(Click on an image for a larger view.)

It doesn’t always happen that way, maybe not even most of the time. Charles Seguin has graphed several names, and in many cases the popularity of the name for boys increases even as the name grow popular for girls.

(Click on an image for a larger view.) 

The lines don’t go in opposite directions, and are often closely parallel, popularity rising and falling for girls and boys and roughly the same time. But in every case — 27 names in all (I did not copy the other two of Seguin’s graphs)  — once the name becomes more frequent for girls, once the blue line crosses to be above the red line, game over. Girls with that name continue to outnumber boys. (Seguin’s paper is here.)

Things may be different in France, at least for one name. Baptiste Coulmont this week tweeted a graph of the name Camille. I know of only three French Camilles, two male — the Impressionist (or is it post-Impressionist?) artist Pissaro and the composer Saint-Saens, both born in the 1830s – and one female, sculptor Camille Claudel, sister of poet Paul Claudel, mistress of Rodin, born in 1864. (I know about her only because I saw the 1988 film with Isabel Adjani.)

Coulmont graphs the ratio of girl Camilles to boy Camilles. Through the first half of the 20th century, the name was twice as popular for boys. Then that relative poularity reverses until, by the turn of this century, there are 15 times as many girl babies given that name. But after 2000, the trend reverses towards boys just as rapidly as it had 30 years earlier for girls. The girl-boy ratio falls from 15:1 to 2:1.

Here is the graph showing frequqencies.

As might be expected, as the popularity of Camille among girls soared, the name lost popularity among boys, falling by 50% over the course of the 1990s. But then came the unusual reversal. As the name lost favor for girls, in rebounded among boys.  Why are French boys returning to the Camille neighborhood as the girls flee? Coulmont does not offer any explanation, only the data. I don’t know enough about current French culture to speculate. For the few other androgynous French names I could find — Dominique, Claude, Yannick — the trends in popularity go in the same directions, separated sometimes by a few years. Camille is unique.

Multiple Negatives and Believable Lies

July 17, 2018
Posted by Jay Livingston

At Language Log (here ), Mark Liberman posted this sentence from a CNN interview with Michael Hayden, CIA director in the Bush 43 administration, about the Mueller investigation.

I would not be surprised
if this were not the last indictment we see
that- that doesn't mention
an American
[emphasis addded]

Does this statement mean that Hayden thinks more Americans will be indicted?

Jake Tapper quickly translated

so in other words there will be another indictment, and you think there'll be Americans involved

Oh those multiple negatives, cancelling each other out. Hayden has three nots.

You have to cut Hayden some slack. He was speaking extemporaneously. But what about writers? I’ve blogged before about problem of multiple negatives in multiple-choice test questions and even the GSS (here).

In today’s New York Times, Mark Landler (here) matches Hayden’s three-in-a-sentence construction. Here’s the second paragraph of Landler’s piece.

Mr. Trump’s declaration that he saw no reason not to believe President Vladimir V. Putin when he said the Russians did not try to fix the 2016 election was extraordinary enough. But it was only one of several statements the likes of which no other president has uttered while on foreign soil. [emphasis added]

I won’t say that Landler’s sentence is not less than incomprehensible. And maybe “Trump said he found Putin’s statement believable” is imprecise and overstates Trump’s credulity. Maybe — but not by much. Here’s what Trump said,

My people came to me, Dan Coats [Director of National Intelligence] came to me and some others, they said they think it's Russia. I have President Putin; he just said it’s not Russia. I will say this: I don’t see any reason why it would be. . .  So I have great confidence in my intelligence people, but I will tell you that President Putin was extremely strong and powerful in his denial today.

Trump does not use multiple negatives. That may be because they pose problems of logic for the speaker, not just the listener. But whatever the reason, this avoidance may be one of the things that fosters the impression that he is a “straight talker.” What he says on a topic may change from one day to the next, but when he voices his view of the day, he states in absolute terms – no reservations, no qualifications.

Double negatives are ambiguous. If we say that someone is “not unfriendly,” we leave open the entire spectrum. from  “possibly somewhat friendly” to “absolutely the friendliest person in the world,” as Trump might put it, especially if he were talking about himself. Trump’s world has no ambiguity. Things that are not good are the worst. Things that are good are the greatest.

Maybe Putin’s denials to Trump about election meddling were similarly uncomplicated — no multiple negatives — allowing Trump to ascribe to Putin’s lies the same credibility that conservatives in the US give to Trump’s lies.

UPDATE:  The press conference happened yesterday. Today Trump issued a clarification that reinforces my point that he doesn’t know how to state ideas involving multiple negatives. In the press conference Monday, on the matter of who was responsible for the hacking and other meddling in the election, Trump said, “ I have President Putin; he just said it’s not Russia. I will say this: I don’t see any reason why it would be.”

Today, Tuesday, Trump corrected himself, reading from a script probably written by Stephen Miller:  “I said the word ‘would’ instead of ‘wouldn’t’. The sentence should have been, ‘I don’t see any reason why it wouldn’t be Russia,’ sort of a double negative,”

He’s probably lying about what he meant to say. But even if he’s telling the truth, he’s saying that the logic of that double negative is a bit too complicated for him, which is why he couldn’t speak it correctly at the time.