Fill in the blanks

Zipf’s Law describes one aspect of the distribution of vocabulary items in a language. It often feels random; it is approximately as random as life.


  1. Given the vocabulary items pictured above, last night I most likely fell asleep listening to the beginning of Céline’s _________________________.
  2. Given the posts that I have been writing lately, I have most recently probably read Queneau’s __________________________________.

French notes

  • le planton: orderly.  In the military sense of the word, not the hospital sense of the word–a hospital orderly is an aide-soignant(e). 
  • l’ambuscade (f.): ambush.
  • en embuscade: in waiting, in ambush.
  • le régiment: regiment.
  • le régiment: military service.  (Native speakers: is this specifically obligatory military service?)
  • l’escadron (m.): squadron.
  • la vérole: smallpox; (slang) syphilis.
  • la décrue: decrease in water level, subsiding of floodwaters.
  • le sapin: pine tree, fir (tree).
  • estompé: (of memory or color) faded, dimmed, dulled
  • le mélo: melodrama.  (Native speakers: is this pejorative?)
  • l’accoutrement: clothes that are amusing.  Get-up.


  1. Voyage au bout de la nuit.  It starts with Bardamu enlisting in the French army during World War I.
  2. Exercices de style, the same story told 99 times in 99 different ways, which I realized I was doing with this series of posts presented as different kinds of tests. 



Instructions: match the vocabulary word with its description. 

  1. Tsipras is pissing off the EU again (don’t pick Athènes, it’s too obvious)
  2. Something you can eat
  3. From an email about my moves on the dance floor
  4. Reminds you of a Queen song

  1. Everything but escarmoucher, guimauve, and épater.
  2. La guimauve (marshmallow)
  3. Épater (haters gonna hate)
  4. Escarmoucher (to spar, physically or verbally)


All questions are to be answered “true” or “false.”  No partial credit.

1. True/False: The guy who was being interviewed on the radio when I came across these words was an Immortal. 

2. True/False: I will have nightmares about these rabbits. 

3. True/False: I need to start leaving less geeky magazines on the coffee table, or no one is going to want to visit me. 


1. True, although to be honest, phagocyter came from an email about what a nice guy I am–none of your goddamned business what the context was.

2. True–you probably didn’t even have to think about this one, right?

3. False–no one is ever going to want to visit me, so why not have cool journals like Natural Language Engineering lying around the house?

Elevators shall not be used in case of fire–seriously??

While I sit on a plane, somewhere past midnight, waiting for a gate to clear…

1. From a linguistic perspective, exactly how useless is this sign?  Give examples. 

2. Multiple choice: The fact that this was my dinner the other night, eaten in bed upon my post-midnight arrival at the hotel, is (a) a classic First-World Problem, (b) just to be expected, (c) better than trying to go to sleep hungry, (d) all of the above. 

3. Knowing that the French verb “to suppurate” (I think it’s actually an inchoative–“to become suppurative,” or something like that) starts  with the sequence ABC makes me feel ______________.

Haussmannian apartment buildings and the zombie apocalypse

Paris will be a great place to survive the zombie apocalypse. Part of the reason why: those stereotypical Parisian apartment buildings.

A recent study evaluated the survivability of cities during the zombie apocalypse.  The approach was rational, modeling a number of factors that contribute to surviving the apocalypse–food supplies, containment, and the like.

Paris was nowhere on the list of best places to be when the mort-vivants come.  This illustrates a problem with the study, because Paris is going to be an excellent place to survive the zombie apocalypse.  In a recent post, we looked at how the history of Parisian street design will make it easier to defend your little home against the marauding flesh-eaters.  However, the next step is important, too–clearing the zombies from the city.  In this, Paris has some distinct advantages.  They come from the design of the stereotypical Parisian apartment buildings.

Picture source:

Part of everyone’s image of Paris is those stereotypical apartment buildings.  They make up about 60% of the city, especially in the older quarters.  You know them: seven stories tall (6 as we count them in France, where the ground floor is the rez de chaussée and the first floor is what we would call the second floor in the US).  The design goes back to the 3rd quarter of the 19th century–roughly 1850 to 1875.  They differ in their details, but the basic design is a 7-story building with horizontal lines of decoration and some kind of additional fancy detail–often a balcony–on the second floor (American 3rd floor, known as the étage noble) and the 6th floor.  The exterior embellishments reflect the original intent of the structure of the floors: shops on the ground level, the shopkeepers on the floor above that, richer people on the floors above that, and the servants under the roof (in the chambres de bonnes).  The picture to the left gives the idea, with more details to be found in the picture below, in French–scroll down past it for the rest of the post, and see the French notes below for a couple of the related vocabulary items.

Picture source:

Second-Empire Paris was a place where there was an enormous amount of money to be made, and an enormous amount of money was made–by people in the “right” social classes.  The Haussmannian apartment buildings played a big role in that money-making.  As Rupert Christiansen puts it in his magnificent Tales of the new Bablyon: Paris in the mid-19th century,

The money was made out of the apartment blocks, those distinctive…six-storey tapered edifices, which have remained the norm in stretches of central Paris until today.  Architectural historians now lovingly catalogue the variations in detail that their iron- and stonework embrace, but the basic construction followed a strict code of regulation: specified height and depth, with an entrance wide enough for carriages and a courtyard for parking, ventilation and fire prevention.

Those apartment blocks are, I think, beautiful–but, their construction (along with that of the wide boulevards that we talked about here) had the effect of displacing an enormous number of people who had previously lived in the old buildings that they replaced.  As Christiansen describes it, Haussmann‘s redesign of  Paris

…remains the nineteenth century’s most radical experiment in shaping and governing urban society… [The logic of the redesign was to] push industry and the sans-culottiste tendency out of the centre of Paris, replacing its warren of dangerous slums with public monuments and commercial development attractive to a new class of clean-living, high-spending, Empire-supporting bourgeois.  In other words, kill off a city built on the virus of poverty and give birth to a financially fecund city resting on clean foundations… [and] keep the construction trade busy and happy, thus preventing a resurgence of the discontents which had led its workers to fan the revolutionary flames in 1848…

You can find more details on what kind of housing the displaced poor ended up in on this web page:  Housing the poor of Paris, 1850-1902.

What’s missing from the study mentioned above will be the relative ease (or difficulty) of   clearing out the zombies afterwards.  The architecture of Paris is made for clearing zombies.  In particular, the typically small apartments limit the number of zombies that you could possibly have to deal with at one time in order to clear a building, as well as reducing the number of rooms from which a zombie could mount a surprise attack–the typical ad for an apartment in Paris (I read them obsessively) is for 30 square meters, versus a more typical 90 square meters in the US.  The vertical design will play a role in this, too–you can clear the zombies one floor at a time, isolating the apartments pretty easily.  A working-class Haussmannian building–if you can find one–is not an imposing edifice (see pictures of mine below, probably built around 1900, a quarter-century after the end of the big Haussmannian boom).  Clearing the zombies out of my building is going to be cake–I look forward to a relatively easy time when the zombie apocalypse comes.


French notes

la mansarde: attic room.  In a Haussmannian apartment building, these are the chambres de bonne–the rooms for the servants.  In theory it’s not legal to rent them out, but of course everyone does, and at the moment there’s a legislative move afoot to legalize those rentals while stiffening the requirements for the modifications that have to be made to them in order to make them moderately habitable.

la lucarne: skylight, dormer.  Look for them in the pictures above.

English notes

to be cake: être fastoche.  This is a shortened form of to be a piece of cake, which means to be fastoche, to be very easy.  How it was used in the post: Clearing the zombies out of my building is going to be cake.  This could also have been put as Clearing the zombies out of my building is going to be a piece of cake.  

mansard roof: An English-language technical term for the type of roof that you see on Second-Empire-style buildings in France and in the US.  As Wikipedia defines the term: A mansard or mansard roof (also called a French roof or curb roof) is a four-sided gambrel-style hip roof characterized by two slopes on each of its sides with the lower slope, punctured by dormer windows, at a steeper angle than the upper.[1][2][3]  I didn’t use it in the post–its presence here is inspired by the definition of the French term la mansarde just above.

dormer: those things with a little roof of their own and a window that you commonly see sticking through the roof of older American houses.  I love them because there is an entire vocabulary for describing them–you can find it, along with plenty of pictures, on this Wikipedia page.


Down the hill comes a bear rug: the non-autonomy of syntax

French people are not rude, and syntax is not autonomous.

I’m walking up the hill to the lab the other day.  It’s a long hill, and I’m old and fat, so I usually stop to catch my breath halfway up.  I’m standing in my usual spot when down the hill comes a man who (I hope) looks even older than I do.  (Down the hill comes an old man…explained in the English notes below.)  Snow-white hair, snow-white beard, and a serious snow-white bear rug rising up out of his coat to merge with his beard.  He and his well-behaved dog walk up to me.  Good morning.  Are you OK?  I take my earphones out.  Oh, yes, thank you–I’m just resting a bit.  You know, there’s a nice path through the forest–the road that you’re going up isn’t really very interesting.  He points out the entrance to the path, but I can’t quite make it out for the trees.  He and his dog walk a bit back down the hill with me and take me right to the beginning of the path.  Go straight for a while–then you can turn left and go down to the town, or turn right and go up to the campus.  My dog likes to walk here–it’s really pretty.  I thank him, I start down the path, and he and his dog head back up the hill.

I get up to “the plateau,” the top of the hill, where the campus is located.  I have only the vaguest idea where I am.  I wander down a little side street, thinking that it might lead to the lab, and pass a construction vehicle.  I hear a voice calling me, and turn back.  Good morning, mister, the construction worker says.  Your earphones are hanging out of your back pocket.  …and he’s right, they are!  I thank him, ask the way to my building, brush the dirt off, and go back to listening to whatever I was listening to.

Google rude french and you will find countless web pages, YouTube videos, etc., on the subject of French rudeness.  Um…seriously??  Not from where I’m standing.  Have I ever had a run-in with a rude lady at the phone company?  Sure–just like in the US.  Have I had enormous numbers of entirely pleasant interactions with  random French strangers?  Sure–just like in the US.  The countries are, in some ways, very, very different–but, the people aren’t any more or less nice in either.  You just have to know how to recognize the very different forms of American and French politeness when you see them–and how to be polite, in both the American and the French ways, yourself.

Here’s a really good explanation of some of the ways that things can go wrong for typical Americans interacting with typical French people in a typical French context–because we mistake French politeness for rudeness.  Note: no one is completely typical.  

English notes

…down the hill comes an old man…: There’s a lot going on here.  I’ll try to unpack it (I think that’s what the kids say these days)–let’s start with looking at the sentence and its context:

I’m walking up the hill to the lab the other day.  It’s a long hill, and I’m old and fat, so I usually stop to catch my breath halfway up.  I’m standing in my usual spot when down the hill comes a man who (I hope) looks even older than I do.

The first thing that you probably notice is the unusual word order–not an old man comes down the hill, but down the hill comes an old man.  This phenomenon goes by a couple of names–a good one is subject-locative inversion.  That is: you are inverting the positions of the subject of the sentence and of a phrase that conveys something that is more or less a location.  (Down the hill is a direction, not a location, but let’s stick with the “simple” name for the moment.)

You probably already understand the sentence, but if you’re not a native speaker, may not be at all clear on when one could use it.  I’m going to draw on this excellent discussion (which itself draws on the Cambridge Grammar of the English Language) and on Wikipedia in trying to explain it.

First of all: as far as I know, you would only use this construction in telling a story.  Much like the passé simple in 20th-century French, this isn’t something that you would use in every-day spoken language.  You can certainly say it out loud, and news reporters often use it to describe a scene when they’re reporting “on location.”  But, you’re not going to stick your head into your wife’s office and say On the table is dinner.  

The circumstances under which you can use this depend on aspects of what we call discourse structure, which you could think of as the way that things that are said form a coherent whole.  We talked the other day about discourse connectives–words and expressions like because, even though, and as a result that establish how the ideas behind two sentences (or whatever) are related to each other.  When talking about subject-locative inversion, there needs to be a particular relationship between the subject (the old man in my sentence) and the locative phrase (down the hill).  The Cambridge Grammar of the English Language specifies that relationship like this:

i. The preposed phrase must not represent information that is less familiar in the discourse than that represented by the postposed NP.

(Preposed means moved to an earlier position.  Postposed means moved to a later position.  An NP is a noun phrase, a group of words that is focussed on a noun.)

Does Cambridge’s claim hold?  Let’s look at the post again.

I’m walking up the hill to the lab the other day.  It’s a long hill, and I’m old and fat, so I usually stop to catch my breath halfway up.  I’m standing in my usual spot when down the hill comes a man who (I hope) looks even older than I do.

We’ve established that there’s a hill involved in whatever it is that I’m going to tell you about: I’m walking up the hill to the lab the other day.  So, in down the hill comes a man, it is indeed the case that the preposed phrase (down the hill) does not represent information that is less “familiar in the discourse” than the postposed noun phrase (a man).  The claim is not falsified by my data.

Back to the Cambridge Grammar of the English Language:

iii. The verb must not represent information that is new to the discourse.

…and back to the post again:

I’m walking up the hill to the lab the other day.  It’s a long hill, and I’m old and fat, so I usually stop to catch my breath halfway up.  I’m standing in my usual spot when down the hill comes a man who (I hope) looks even older than I do.

Is it, in fact, the case that the verb does not represent information that is new to “the discourse?”  Actually, I don’t know how you would evaluate that, one way or the other.  Certainly the whole “discourse” up to that point has been about my movement, so I guess you could say that the verb to come is not “new”–it’s in the same lexical field.  However, if you accept that analysis, then unless you insist that the verb be identical to some preceding one, I don’t know how you could demonstrate that the verb did represent information that’s “new to the discourse.”

(You might have noticed that my quotes from the Cambridge Grammar of the English Language went straight from (i) to (iii).  This is because I don’t know what the hell (ii) is talking about.  See this page if you want to see it for yourself.)

This kind of phenomenon turns out to be important in one of the deep theoretical issues in linguistics.  For various and sundry reasons, Chomsky and his ilk would like to posit what they call “the autonomy of syntax”–the idea that syntax is a part of human linguistic knowledge (the kind that we all have unconsciously, not the kind that people like me get in graduate school) is “autonomous”–that is, that it operates independently of the rest of language.  When you have a situation like we see with subject-locative inversion, where the syntax is dependent on something else–in this case, on the structure of the discourse, or more specifically, whether something is “old information” or “new information”–it makes it tough to argue that syntax is autonomous.

The best place to survive the zombie apocalypse, and why Parisian streets are wide

Why are Parisian streets wide? They mostly aren’t, and that has everything to do with surviving the zombie apocalypse.

Boulevard Montmartre by Camille Pissarro, 1897. Picture source:

I recently read a news story on the subject of the best and worst cities for surviving the zombie apocalypse.  The model took some reasonable things into account–ability to maintain a food supply, ease of containment, stuff like that.  Scores ranged from a low of 8.69 (New York, at 53rd place) to 48.33 (Boston, in first place).  (You have to know the range of a number to have the faintest clue how to interpret it.)  However, ultimately, the model is bullshit.

Here’s the thing: first of all, the study only looked at US cities.  Excuse me: most people in the world are not in the US, and if the US hopes to survive the zombie apocalypse, it’d better hope like hell that the rest of us do, too.  Second of all: most of the things in the model had to do with the ability of the city to deal with the virus, while none of them had to do with the ability of the city to deal with the zombies.  When you introduce the basic physical geography of the city itself into the model, Paris starts looking really good, for the following reason: our many tiny streets will be easy to keep the zombies out of, because they present relatively small openings at each end.  A Special Forces friend once told me this: the purpose of a wall is so that you can shoot people as they come around it.  The tiny openings on our tiny streets will make them easy to barricade, so it will be easy to nail the zombies as they try to get past the barricades.

Now, I know what you’re thinking: Paris is all wide, tree-lined boulevards, right?  Some of it is, sure.  Most of it isn’t, though, and this is quite related to why those of us in Paris are going to be in a good position when the zombie apocalypse comes.

Part of many people’s stereotype of Paris is the grands boulevards, the wide streets that characterize much of the older part of the city.  Indeed, they are for real: as Wikipedia puts it,

The boulevards…now define Paris, with uniform façades and overhanging balconies stretching along them. These are immediately recognisable, and are under the strict control of Paris’ urban planners.

It’s not clear that they’re just there to be pretty, though.  A major part of the enormous urban renewal project that took place in Second Empire Paris between about 1850 and 1875 or so was the tearing down of the buildings around the old twisting, crooked medieval streets, and replacing them with these broad boulevards.  A number of reasons have been proposed for this particular aspect of the project, most of them not what you would expect.

  • Theory: the streets were widened to make it harder to barricade them in times of political unrest.  From Wikipedia: The center of the city was also a cradle of discontent and revolution; between 1830 and 1848, seven armed uprisings and revolts had broken out in the centre of Paris… The residents of these neighborhoods had taken up paving stones and blocked the narrow streets with barricades, and had to be dislodged by the army.[6]
  • Theory: the streets were designed to be wide enough to permit a cavalry charge in times of political unrest.  As a web site for buying property in Paris puts it: Hoping to ease the movement of the cavalry if another conflict arose, as well as considering the dimensions of the steering locks for the horse drawn artillery, Napoleon III carefully calculated the width of the avenues based on these factors, as well as aesthetic ones.
  • Theory: Haussmann realized that the city would be able to function only if its populace and its supplies could be moved around quickly.  Wikipedia describes the pre-Haussmann situation like this: Traffic circulation was [a] major problem. The widest streets in [some neighborhoods] were only five meters wide; the narrowest were only one or two meters wide.[4] Wagons, carriages and carts could barely move through the streets.[5] 

Haussmann had to deal with this impediment if the redesigned city were to work.  As Rupert Christiansen puts it in his truly magnificent Tales of the New Bablyon: Paris in the mid-19th century, which tells the story of the Paris Commune, the siege of Paris during the Franco-Prussian war that preceded it, and the tail of the Second Empire, which preceded the siege:

Haussmann understood with genius one profound necessity of the age: the modern city’s dependence on its lines of communication, its capacity to move at speed.  [A] cliché is the notion that the chief reason for planning wide, straight boulevards was to allow for the strategic movement of troops in the event of rioting and to prevent the throwing up of barricades, but there is no evidence to suggest that this was anything more than a minor incidental consideration…The absolute priority was moving from A to B with the minimum amount of interference, in the minimum amount of time.

Recall that in Second Empire Paris, the economy was booming and there were these brand-new railroad termini.  If you wanted to be able to sell stuff, you had to be able to get it into town, and if you wanted to get that stuff into town, the other guy had to get his train emptied quickly so that you could get your train in there.

In fact, there was every reason to worry about barricading the streets of Paris.  Note that Christiansen doesn’t say that keeping rebellious Frenchmen from controlling the streets wasn’t an issue–he says that it was nothing more than “a minor consideration.”  Why worry?  In the revolution of 1830, the rebelling Parisians (it’s almost always Parisians that rebel, or at least that start it–the Fronde is the only exception that I can think of.  Native historians?)  immediately starting throwing up barricades.  A quote from the Wikipedia entry on the 1830 revolution:

In only a day and a night over 4,000 barricades had been thrown up throughout the city.

In the revolution of 1848, barricades were, again, a crucial part of the kick-off of the Parisians’ rebellion.  From the Wikipedia page on the revolution in question:

On 23 June 1848, the working class of Paris rose in protest over the closure of the National Workshops. On that day 170,000 citizens of Paris came out into the streets to erect barricades.[25]

They were effective, too, those barricades:

To meet this challenge, the government appointed General Louis Eugène Cavaignac to lead the military forces suppressing the uprising of the working classes…. Cavaignac’s forces started out on 23 June 1848 with an army composed of from 20,000 to 30,000 soldiers of the Paris garrison of the French Army.[25] Cavaignac began a systematic assault against the revolutionary Parisian citizenry, targeting the blockaded areas of the city.[26] However, he was not able to break the stiff opposition put up by the armed workers on the barricades on 23 June 1848….Cavaignac’s forces were reinforced with another 20,000–25,000 soldiers from the mobile guard, some additional 60,000 to 80,000 from the national guard.[27] Even with this force of 120,000 to 125,000 soldiers, Cavaignac still required two days to complete the suppression of the working-class uprising.

What do you make barricades of?  There are certainly a number of options.  Furniture.  Trees.  In 1848, omnibuses were used.  But, mostly, you build them out of paving stones.

Note several places in the photograph where you can see the edge between where the stones are still in place and where they’ve been pulled up to build the barricade.  Original caption: avril 1871. Coin de la place de l’Hôtel de Ville et de la rue de Rivoli. Picture source:

Here’s a quote from a recounting of the beginning of the 1968 riots in the Latin Quarter:

The police surrounded the march on the Boulevard Saint Michel, halting the demonstrators’ progress. The students decided to occupy and defend the Latin Quarter, they spread into neighbouring streets and quickly started constructing barricades out of anything they could find. Many sympathisers – young and old – flocked to the Latin Quarter to join in the student revolt after listening to their exploits on independent radio stations. A passing builder demonstrated the use of a pneumatic-drill to students who were trying to dig up the cobblestones (the soon-to-be infamous pavées) from the street to add to their arsenal. (bolding by me)

Notice the bolded text: the students started on Boulevard St-Michel, and then when they wanted to barricade the place, they went off into the side streets.  A small detail, but a telling one.  There’s a reason that you would need to leave the Boulevard Saint Michel if you wanted to build barricades–the Boulevard Saint Michel is a classically wide Haussmannian boulevard, and you’d have a hell of a time trying to block it off.  In contrast, the side streets in the Latin Quarter are stereotypically narrow.

And barricades are not all that pavés, the traditional Parisian cobblestones, are good for: they are “traditionally” used for throwing at whomever you happen to be barricading the streets against.  This goes way back–here’s a quote about paving stones being used as weapons in the 1830 revolution: Paving stones, roof tiles, and flowerpots from the upper windows… began to rain down on the soldiers in the streets”.[10] 

Post-riot rubble, Paris, May 1968. Picture source:

…and it was a serious activity during the 1968 student riots.

So, yeah, Paris has some grands boulevards, but most of it is small streets, and it’s going to be a good place to be if you’re interested in keeping your street clear of zombies during the apocalypse.  Un homme averti en vaut deux: a word to the wise is sufficient.

Want more information about streets in Paris?  Here are a couple miscellaneous links.  French and English notes below.

French notes

“Queue de paon” (peacock tail) paving stone pattern. Picture source:

la queue de paon: peacock tail.  This is the term for a particular pattern of paving stones that you’ll sometimes see in Paris.

English notes

cobblestone: A stone used for paving a street.  Merriam-Webster defines it more generally: a naturally rounded stone larger than a pebble and smaller than a boulder; especially :  such a stone used in paving a street or in construction.  The word goes back to the Middle English period (which was roughly 1100-1400).  In restless dreams I walked alone // Narrow streets of cobblestone (Paul Simon and Art Garfunkel)  Example from WikipediaCobblestones are typically either set in sand or similar material, or are bound together with mortar.  How it was used in the post:  And barricades are not all that pavés, the traditional Parisian cobblestones, are good for: they are “traditionally” used for throwing at whomever you happen to be barricading the streets against.

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