To the store went Curious George

Curious George.  Picture source:
Curious George. Picture source:

I just sent this email to a coworker: The data sets are linked to from this article.  If you’re a native speaker of English, that sentence probably doesn’t look particularly unusual, and you probably agree that it could be paraphrased as There are some data sets.  There are links to the data sets.  The links are from. 

If you’re not a native speaker, though, you might well have stumbled on this sentence due to the sequence to from:

The data sets are linked to from this article.

How can to from possibly be meaningful?  It works because of one of the most important observations in the history of linguistics.  In a 1957 book, Verbal Behavior, the famous psychologist B.F. Skinner, the father of behaviorism, maintained that the syntax of English could be described by something called a finite state machine.  Finite state machines basically describe the possible sequences of something–words, in this case.  In a review that he wrote in 1959, Noam Chomsky poked a huge hole in Skinner’s claim.  He made the point that syntax is not about sequences, per se, at all–rather, it’s about structures.  The data sets are linked to from this article works because the sequence to from doesn’t figure into how that sentence is produced or interpreted.  Rather, the sentence is composed of a set of structures: [[The data sets] [[are [linked to] [from [this article]]]].  The sentence is not a sequence of words; to the extent that it’s a sequence of anything, it’s a sequence of phrases (syntactic structures), and it’s really not so much about the sequence of anything as it is about structure.

Within that sentence, linked to is what most people would call an idiom.  Rather than a preposition, it contains what English-language linguists call a particle.  (To see how it’s different from a preposition, consider these pairs of sentences:

  • Curious George went to the store.
  • To the store went Curious George.

No problem–both of those are perfectly acceptable English.  Now think about these:

  • Curious George linked to the blog post.
  •  * To the blog post linked Curious George.

The asterisk (*) is the linguist’s way of indicating that * To the blog post linked Curious George is completely unacceptable–you just can’t say that in English.  Why?  It’s because in go to, the to is a preposition.  In link to, it is a particle.  Particles and prepositions behave differently.  One of the ways in which they behave differently is that prepositions and their objects have some ability to move away from their verb, as in To the store went Curious George.  Particles can’t do that.  So, the native speaker of English has no problem with the to from in The data sets are linked to from this article–there really isn’t any other way that the native speaker can analyze it.  If you want more information on what I’ve called idioms (verb + particle), as well as other English constructions that use particles, see this article on the Linguistics Girl blog.

Chomsky went on to become the most important linguist of the 20th century.  Not the best linguist of the 20th century–but, unquestionably the most important linguist of the 20th century.  He redefined what the field was about, in almost every respect.  Remember this cartoon from a couple days ago?  It was about him.  I can tell you that only a vanishingly small number of linguists are the subjects of cartoons—we just don’t typically get very famous.  “Movement phenomena” like the examples that we saw in this post were a big part of the motivation for his theories, and there are movement phenomena in French, as well as in English.  We’ll see some movement phenomena in the formation of various types of questions, in particular.  But, more on that later–time for dinner.  For now, let’s just remind ourselves that the Curious George books were originally written in French.

Labiodentals: lips and teeth

Picture source:
Picture source:

In hopes of getting me able to pass the DALF, we’ve been looking at verbs whose infinitive ends with IR.  We’ve looked at regular IR verbs, like finir.
We’ve looked at irregular verbs like courir, dormir, partir, sentir, and sortir:

finir courir dormir partir sortir
je finis cours dors pars sors
tu finis cours dors pars sors
on finit court dort part sort
nous finissons courons dormons partons sors
vous finissez courez dormez partez sortez
ils/elles finissent courent dorment partent sortent

Now let’s look at yet another set of irregular IR verbs. This time, we’re going to figure out a way to remember which verbs belong to this class:

ouvrir couvrir offrir souffrir
je ouvre couvre offre souffre
tu ouvres couvres offres souffres
on ouvre couvre offre souffre
nous ouvrons couvrons offrons souffrons
vous ouvrez couvrez offrez souffrez
ils/elles ouvrent couvrent offrent souffrent

First, what’s unusual about this class?  It’s the endings–they are the same as for ER verbs, which the IR verb endings usually differ from quite a bit.

Now, how can we remember these?  It turns out that finding patterns in this kind of data is what linguists do all day.  Here, the pattern appears to be related to the consonantal structure of the verb roots.  To understand what’s going on, you need to know a few things about the sounds of language.

  1. To simplify quite a bit: speech sounds are produced by making air leave the lungs through the mouth.  Look in the mirror while you make the sound ah.  Close your mouth and try to make the sound ah.  Doesn’t work.
  2. Consonants are made by obstructing the flow of air through the mouth.  Look in the mirror while you say bah-bah-bah.  See how the flow of air is obstructed completely when you make the sound that we represent with the letter b?
  3. Different consonants are made by obstructing the flow of air at different places.  Look in the mirror while you make the sounds bah and kah.  Does your mouth look the same, or different?

Now that you have some of the basics of how consonants work, let’s look at the sounds in the roots of these verbs.  You notice that (1) there is a cluster of consonants (i.e., more than one); the second consonant is r; and the first consonant is one of f or v.  Look in the mirror while you make the sounds fah and vah.  You’ll notice that the obstruction for these consonants is made with the lips and teeth.  (We’ll talk about how you differ between the f and the v some other time.)  Consonants that are produced with an obstruction made by the lips and teeth (actually, just one lip) are called labiodental consonants, or labiodentals. 

It turns out that French has exactly two labiodentals: the sound that we usually spell with the letter f, and the sound that we usually spell with the letter v.  (In the languages of the world, there are two other labiodental consonants.  We don’t have a way to spell them in English, but you probably have one of them–the labiodental nasal–in the English word emphasis.)  So, we can say that you see this particular pattern of verb endings when a verb has a consonant cluster in the middle, the first consonant is a labiodental, and the second consonant is r. 

I know of three verbs that don’t have the labiodental-r root but that do have this pattern in the present indicative.  However, they behave differently from the others with respect to their past participles.  Those verbs are cueillir (to pick, to gather), acceuillir (to receive, to greet), and recueillir (to collect, to gather).  The present indicative is like the labiodental-r verbs that we’ve seen.

However, the past participle of these verbs is different from the labiodental-r verbs: couvert versus ceuilli.  There are also a couple of verbs that have -aillir in the stem that have the same present tense pattern, maybe–one of my Bescherelles has a note about how even famous writers sometimes don’t follow this pattern for those verbs.

So, with a little bit of linguistics, you don’t have to memorize the fact that it’s verbs with fr and vr in the root that have this conjugation–all you have to do is remember that it’s verbs with a consonant cluster in the root where the first consonant in the cluster is a labiodental.  Fun, huh?

Postmodernism, Burkina Faso, and dissolving stuff

"Can Coke dissolve things?"  Picture source:
“Can Coke dissolve things?” Picture source:

It’s sort of a postmodern cliché that the media exerts tremendous influence over what, and how, we think.  I didn’t take this seriously until I was in Paris last summer.  We went through a period of pretty serious anti-Semitic violence–a synagogue was attacked by a mob during services, Jewish businesses were bombed, cars in a Jewish neighborhood were set afire.  A young woman was raped.  This was reported on my favorite American news channel as follows: “There was unrest in Paris yesterday.”  Really?!  To my astonishment, the most reliable source of news about what was going on turned out to be Twitter.

I thought again about how different media have different takes on what is newsworthy during the recent coup in Burkina Faso.  There wasn’t much news coverage of it in the United States, but it was covered quite heavily in the French media.  In fact, Radio France International’s podcast about French words used it to establish the context for a discussion of the word dissoudre, “to dissolve,” as in the rebels dissolving the government.

Dissoudre turns out to fit nicely into our recent discussions of verb conjugations.  In general, verbs ending with -re tend to be at least somewhat irregular.  In that way, dissoudre is a doozy.  Let’s look at the present tense:

je dissous nous dissolvons
tu dissous vous dissolvez
il/elle/on dissout ils/elles dissolvent

Where does the LV come from? Where does it go? The mysteries of morphology. Here are some verbs that are conjugated like dissoudre.  These might be the only three that follow this pattern; they share other oddities, including irregular past participles and possibly not having passé simple or imperfect subjunctive forms.

  • absoudre: to absolve.
  • résoudre: to solve, resolve.

A magot is not a maggot

Simone de Beauvoir in an unnamed cafe. Photo source:
Simone de Beauvoir in an unnamed cafe. Photo source:

I’ve always been attracted to depth over breadth.  Visiting a new city, I might eat in the same restaurant every night, trying to explore the entire menu.  Discovering a new bookstore, I’m unlikely to only visit the sections that I’m interested in and know that I’ll buy from, but very likely to walk around the entire store, seeing what it’s strong in (in Denver, the former Borders in Northglenn for judo, the Tattered Cover for history, Barnes and Noble for obscure French books; in Paris, Gibert Jeune for theoretical semantics, Attika for bilingual French/English novels, and Gibert Joseph for cheese, kids’ books, and general linguistics).  This makes me a super-boring person to visit a new city–or a new bookstore–with, but I love the feeling of knowing a place deeply, and prefer it to knowing lots of places more shallowly.

I’d always been interested in an in-depth exploration of the cafes of Paris, but had trouble finding a good, and preferably literary, guide to them.  Then today I read a chapter in John Baxter’s Five nights in Paris: After dark in the City of Lights on just that topic.  Baxter starts by debunking the myth that Parisian writers work in cafes.  He says that Hemingway was the last writer of importance to actually write in cafes in Paris.  He also, however, quotes a friend as saying that only one of her writer friends still writes in a Parisian cafe.  There’s a clear inference from this: writing in Paris cafes has certainly been popular, and people still do it.  I will even confess to having spent a pleasant afternoon or two sitting on the terrasse of a Paris cafe working on a (not very interesting, unfortunately) book myself.  He lists a number of cafes frequented by writers, musicians, or other folks of interest; I’ll round it out a bit with material from Graham Robb’s Parisians: An adventure history of Paris.

  • Cafe Procope: I’ve heard different stories as to whether this was the first cafe in Paris, or the first successful cafe in Paris.  Graham Robb describes the original offerings: coffee and sherberts.  Habitués over the years have included Voltaire, Rousseau, Benjamin Franklin, John Paul Jones, Thomas Jefferson, and Léon Gambetta.  6th arrondissement, which I think is not its original location, but I haven’t been able to verify that.
  • Les Deux Magots: one of the most famous cafes in Paris.  It was opened in 1885 in the former location of a silk store.  According to the web site, the name comes from the two magot, or seated Oriental figurines, that are mounted on the wall and are the last vestiges of the silk shop.  Expect to pay about double the cost of a coffee or beer anywhere else in town for the privilege of hanging out in the same place as Rimbaud, Mallarmé, André Breton, Paul Eluard, Simone de Beauvoir, Jean-Paul Sartre, Ernest Hemingway, Albert Camus, Pablo Picasso, James Joyce, and Bertolt Brecht, once did (list from Wikipedia and the web site).  I was really surprised at how small it was on the inside–it’s easy to imagine the air choking with cigarette smoke and the smell of espresso.  6th arrondissement.
  • Cafe Flore: right up there with Les Deux Magots in the fame department, and still popular.  Like Les Deux Magots, it was built in the 1880s.  Habitués have included Pablo Picasso and Zhou EnlaiAdam Gopnik‘s essay A tale of two cafes (reprinted in his Paris to the Moon, which I can’t recommend highly enough–my favorite book about Paris) contains a variety of wildly speculative explanations for why Les Deux Magots stopped being the cool place to be while the Flore remained popular; all or none of them may be true.  6th arrondissement.
  • Cafe Beaubourg: Edmund White hung out here.  As Baxter relates, and White himself says in his memoir Inside a Pearl: My years in Paris, part of the attraction was walking by the air intake of the Institut de Recherche et Coordination Acoustique/Musique afterwards, so that his dog could poop in it–a dispute over being refused an interview.  4th arrondissement, in the Marais.
  • La Closerie des Lilas: Hemingway liked to write here.  Other habitués have included Pablo Picasso, Jean-Paul Sartre, André Gide, Paul Éluard, Oscar Wilde, Samuel Beckett, Man Ray, Ezra Pound (list from French Wikipedia page).  6th arrondissement.
  • Café Fleurus:  I haven’t been able to find much in the way of information about this one, but it appears that Gertrude Stein lived right up the street, at 27 rue de Fleurus.  14th arrondissement.
  • Wepler: This was a favorite of Henry Miller.  The character Joey in Miller’s book Quiet Days in Clichy has a relationship with a prostitute that he meets there.  18th arrondissement.
  • Tournon: This was a favorite of post-war African-American expats in Paris, as well as the early male-female transsexual April Ashley.  6th arrondissement.

As Baxter points out, these days Paris cafes are about food as least as much as coffee, but I wouldn’t necessarily recommend eating at any of these.  A number of the cafes that I’ve listed are in the 6th arrondissement.  If a cafe tour takes you there, I would recommend that you eat at Au Pied de Fouet.  This is an island of affordable meals in the midst of a sea of super-overpriced places, and the food is some of my favorite stuff in Paris, particularly for French stuff (yes, Paris has a bazillion great restaurants serving things other than French food).

For a Zipf’s Law connection, let’s look at the vocabulary items that appear in some of the cafe names:

Do you have a cafe to add to my list?  How about mentioning it in the Comments?  I’ve focused here on cafes with some literary history; you can find a list of other cafes here.  Some of them look pretty interesting.

Irregular IR-class verbs, or why I’m not losing weight

Exploring group III French verbs through my failure to lose weight.

So embarrassing–I had a great opportunity to use an obscure IR-class verb (mincir, meaning to lose weight or to make you look thin) yesterday, but in the first person singular present tense, not the third person plural present tense that we worked on last time–and I blew it.  Attempting to explain the connection between the delicious French-Canadian dish called poutine–fries covered with cheese curds and gravy–and why I’m not losing weight, I conjugated it as an ER verb, not IR.  !@#$%$!  I guess I just gotta work on those IR-class verbs some more.  So, for the moment let’s just point out that there’s a Montréal restaurant, La Banquise, that serves 25 different kinds of poutine, agree amongst ourselves that I’ll go there this week when I visit our neighbors under the Big Red Maple Leaf, and focus on irregular IR verbs.

In that spirit, let’s look at the present tense of some irregular IR verbs.  In the singular forms, the final written consonant is the same, but where the regular IR verbs have the vowel i in front of that consonant, the irregular IR verbs do not. We’ll use finir (to finish) as our prototype of a regular IR verb–all of the other verbs in these tables are irregular IR verbs:

finir courir dormir partir sortir
je finis cours dors pars sors
tu finis cours dors pars sors
on finit court dort part sort

In the plural forms, the regular IR verbs (like finir) and the irregular verbs (all of the other verbs in this post) are quite different, and actually look a lot like ER verbs:

finir courir dormir partir sortir
nous finissons courons dormons partons sors
vous finissez courez dormez partez sortez
ils/elles finissent courent dorment partent sortent

Similar verbs include mentir (to lie), sentir (to smell), and compounds of all of these.

How many verbs like this are there? It’s surprisingly difficult to say. It’s even unclear what exactly “this” means. The traditional answer would be “the set of third-conjugation verbs,” but “third-conjugation verbs” include a number of verbs of entirely different classes. Just looking at the example verbs on this page, there’s a clear difference between verbs like courir and verbs like dormir–they share the same endings, sure, but the stem of the verbs like dormir lose a consonant in the singular forms.  Would you count mourir?  The endings are the same, but there’s a change in the stem vowel.  How about démentir (to deny)?  It’s conjugated like mentir,  but while the past participle menti is invariable, the past participle of démentir can be inflected for gender, and be démenti or démentie.  Does it count as like “this”?  And, there are words related to the words that I’ve used as examples here.  For example, related to courir (to run), we have (from the web site L’Obs–la conjugaison):


Counting word types is always an ugly business–this shows you one thing that contributes to that kind of ugliness.  Mincir (to lose weight) is totally regular, by the way, although at this point in my life, for me to lose weight would, unfortunately, be quite irregular.

3rd person plural present tense of regular IR-class verbs: anonymous sex in the Tuileries Garden

The Tuileries Garden today. Source:
The Tuileries Garden today. Source: “Tuileries gardenview” by User:Munford – Own work (Taken by me). Licensed under CC BY-SA 3.0 via Commons –

There’s good evidence that men have been jumping the walls of the Tuileries Garden (built in 1564) to have sex with each other at night since at least the 17th century.  We know about this because although France has no laws forbidding sex or marriage between consenting adults today, it wasn’t always so enlightened.  At one time, the police paid men called mouches (flies) to entrap gay men into making sexual propositions and then have them arrested.  The police reports of the mouches give quite a bit of insight into gay cruising culture in the decades before the French Revolution (1789).  In 1791, the new French penal code decriminalized homosexuality.

Now that I’ve got your attention, it’s time to get back to the basics.  I’m preparing for a French certification test (see this post for a description of the oral comprehension portion), and I am realizing that I am woefully out of practice with the conjugations of some verb classes.  About 20% of French verbs end with -ir.  The 3rd person plural present tense of these verbs (they walk, they are walking, etc.) is a weakness for me, so humor me and let’s work on it.

For the 80% or so of French verbs that belong to the ER class, the 3rd person plural present tense is pronounced the same as the 1st, 2nd, and 3rd person singular.  For the IR verbs, the stem is pronounced the same as the 1st and 2nd person plurals.  (It’s always written differently from all of the others.)  The ending for the 3rd person plural present tense of regular IR verbs is -issent:

je finis nous finissons
tu finis vous finissez
on finit ils finissent

With that reference in hand, let’s practice. A list of sentences follows. Change the highlighted pronoun and verb to ils, and give the appropriate verb form.

  1. Nous ne choisissons pas notre orientation sexuelle.  We don’t choose our sexual orientation.
  2. Je vomis les “mouches.”  I loathe the “mouches.”
  3. Je suis ravi qu’il abolit les lois contre l’homosexualité.  I’m delighted that he is abolishing the laws against homosexuality.
  4. Est-ce que tu rougis quand je parle de ces affaires?  Do you blush when I talk about these things?
  5. Réfléchissez-vous à ce que j’ai dit?  Are you thinking about what I said?


  1. Ils ne choisissent pas leur orientation sexuelle.
  2. Ils vomissent les mouches.  They loathe the mouches.
  3. Je suis ravi que qu’ils abolissent les lois contre l’homosexualité.
  4. Est-ce que ils rougissent quand je parle de ces affaires?
  5. Réfléchissent-ils à ce que j’ai dit?
Curative Power of Medical Data

JCDL 2020 Workshop on Biomedical Natural Language Processing


Criminal Curiosities


Biomedical natural language processing

Mostly Mammoths

but other things that fascinate me, too


Adventures in natural history collections

Our French Oasis


ACL 2017

PC Chairs Blog

Abby Mullen

A site about history and life

EFL Notes

Random commentary on teaching English as a foreign language

Natural Language Processing

Université Paris-Centrale, Spring 2017

Speak Out in Spanish!

living and loving language




Exploring and venting about quantitative issues