Last night, having just watched a documentary on the Connecticut River flood of 1936, my wife and I discovered that our basement had flooded — apparently the sump pump had failed. So this morning we called the Barstows (it’s great to have contractors you can rely on in emergencies) and they sent a crew over within half an hour, unclogged the pump (“you should have it checked every year”), vacuumed the floor, and left. I investigated and discovered that, although most of the boxes were up out of harm’s way (a precaution we took after the last flood, a decade or so ago), there was a box of sf books and magazines that had gotten wet, so I brought it up, opened it, and set everything out to dry. Part of the contents was a set of Worlds of Tomorrow, a companion magazine to If which I was buying in the mid-’60s; I opened the January 1966 issue at random and found in the table of contents “How to Understand Aliens,” by Robert M. W. Dixon. “The Robert M. W. Dixon?” I thought, and sure enough, the Australianist linguist about whom I posted repeatedly in January 2006 (1, 2, 3, 4, 5, 6) had written both stories and articles for sf magazines back in those days.
I wish I could send you to an online version of “How to Understand Aliens,” but there doesn’t seem to be one. You can see the beginning in accursed snippet view here, but that’s not much help. You can read the article here (thanks, Owlmirror!). I’ll copy out a short passage that will give you some idea; the whole thing is well done, as one would expect, and hopefully gave some readers (and writers) a better idea of how language works:
Space linguists could gain valuable practice in unravelling bizarre languages by having a preliminary workout on a terrestrial language before venturing into extra-terrestrial contact. For instance, a cadet linguist thrown amongst a tribe of Australian Aborigines would be able to get an idea of the variety of similar meanings a word can have when he learned that gargal could mean, firstly, the upper part of the human arm as it meets the body; secondly, the lower part of a branch, where it meets the trunk of a tree; and thirdly, the mouth of a stream where it flows into a larger river. And of how the meaning of a word can be extended to apply to new situations: the word for ‘hollow log’, maralu, being taken over to apply to ‘shirt’ when the aborigines first came into contact with white men wearing this novel garment.
Hello maralu
Goodbye heart.
Why do linguists like science fiction?
I hate basements. If you want more space, build an extra floor above ground.
an extra floor above ground
What about meteorites? Or space junk coming down, for that matter? Or raccoons? No matter what you do, shit/s gonna happen!
I like shit to happen in broad daylight. Basementwise, raccoons are moot.
Something that just occurred to me: why weren’t the Barstows locked down?
Fixing flooded basements is an essential service if ever there was one.
Any self-respecting space alien is going to be a whole lot aliener than an indigenous Australian. (Possibly the one point on which Chomsky is definitely right.)
I mean, we’re practically talking about French here.
Talking of which, there are far too few Frenchmen in Space in science fiction.
The superb Fifth Head of Cerberus has some.
And a very well-known writer has some in an excellent book of his which I keep trying to get people to read so that I can bang on about how wonderful it is; unfortunately, it’s impossible to explain how wonderful it is without major spoilers, and now that I’ve said it involves Frenchmen in Space I can’t even name the author without spoilers. It’s so frustrating.
Jean-Luc Picard doesn’t count. Obviously. Nor Valerian (even though he actually is French.)
Well, if you don’t mean J. Verne I’ll eat a broom (Ger.loc.).
So much more interesting than the silly movies that have been made from his books. Full of ideas and human o’erweenieness.
In re Cerberus:
These works are set on two colony worlds, 20 light-years from Earth, the double planets of Sainte Anne and Sainte Croix, originally settled by French-speaking colonists, but lost by them in a war with an unnamed enemy.
I thought they are their own worst enemy, languagewise ?!
Any self-respecting space alien is going to be a whole lot aliener than an indigenous Australian.
As Dixon says; the concluding sentence: “However much we may speculate about alien languages, they are certain to be different from anything we know or could, without encountering them, imagine.”
Preach it, Bob!
I didn’t mean Jules Verne, on the grounds that he’s a Frenchman about Space, rather than in Space.
The Fifth Element ? Hmm, no, that’s not a book.
Alain Damasio ?
Also, although Bruce Willis has many virtues, being French is not among them.
https://archive.org/details/Worlds_of_Tomorrow_v03n05_1966-01_dtsg0318.Anon/page/n113/mode/2up
It’s not that there aren’t French science fiction writers: it’s the dearth of actual Frenchmen in Space in science fiction that I lament. If we cannot encompass a Galactic Frenchman imaginatively …
Thanks, Owlmirror! I’ll add it to the post.
David E, you must be a devotee of the Times crossword puzzle. J’accuse. “Galactic Frenchman” is one of those clues.
@Owlmirror:
Many thanks. That’s an interesting article indeed.
I note (with enthusiastic approbation) on the last page RMWD’s rubbishing by implication of the Chomskyan notion that you can divorce the study of syntax from meaning, and, with equal enthusiasm, his basic point that you can’t understand the meaning without understanding the culture. Preach it, Bob!
His space linguists seem to have been faced with a remarkably easy challenge though. Thinking about just how difficult it might have been made me remember Nagel and the Bat:
https://warwick.ac.uk/fac/cross_fac/iatl/study/ugmodules/humananimalstudies/lectures/32/nagel_bat.pdf
I was mocked by a science teacher in high school for reading trashy sf magazines, but I was learning as much from them as I was in class. Dammit.
As far as I know (not much) space experts in language (aka linguists) are charged with trying to understand alien’s language. For some reason no one seems to think that we need a specialist to explain to the aliens our language (to assuage David Eddyshaw, let it be French). Should we send a teacher into the space instead?
Reminded me of
https://languagelog.ldc.upenn.edu/nll/?p=29480
Actually, I think a difficulty with D.O.’s very reasonable suggestion is that we ourselves don’t understand how our own languages work all that well.
The aliens would probably be better off ignoring everything that we say about our languages and concentrating on analysing what we actually say and how (and when) we say it (advice given to many a terrestrial field linguist, nowadays somewhat tempered by political correctness in the way it’s actually expressed, but still pretty good advice.) They’d probably notice things that no human being ever would, too (and perhaps we would be able to return the favour.)
I used to shit in broad daylight, but it’s been many years since I was that far back in the woods. Still, accidents happen.
But as for basements, if you have to dig multiple meters down to get below the frost line, you might as well put down a concrete slab and a stairway of some sort.
If ever aliens come visiting, it may well be that they speak some space lingua franca that has been optimised for being easy to learn across intelligent life forms. Our resident space lizard should know…
So you’re not the Pope, I had suspected as much.
Well the frost line in..let’s say Yonkers, is three feet six inches. Even if New Yorkers have shrunk since my time you’d need more than twice that height for a basement you could legally occupy. And moving all that dirt is an awful lot more work than moving the air out of the way and adding an extra floor. Concrete slabs are awful, they’ve no bounce, I like wood.
That said, some people are able to imagine a lot:
“It’s perfectly possible for a language to have a grammar whose sentences are structurally incompatible with a representation in terms of tree structure, and here’s an example to prove it.”
I was speaking of my house further north and further inland, which is set into the slope of a hill. As a result, what is basement on the west side of the house is ground level on the east side, and the living floor (and its exterior extension, the deck) is three steps above ground level on the west side and, say, 15 steps on the east side (I haven’t counted them lately).
It’s the east side that’s along the road, so you can go into the (ground-level) basement door and up the inside stairs, or up the outside stairs and into the living-floor door; the two stairways are adjacent but separated by the north wall. (I hope this makes sense: plan views would be worth a lot more words than this.)
I see that the frost line in New York State generally is 4 feet unless overridden by local ordinance. When we were building a cabin for my older sister on the same property, it was situated on four concrete pillars each about five feet long, but with their tops at different heights above the ground (again, on a hill, it’s all hills there).
=====
Actually, non-tree-structured languages are no further away than Switzerland. As Shebs 1985 points out, Swiss German Jan säit das mer em Hans es huus hälfed aastriche ‘Jan says that we the-DAT Hans the-ACC house helped paint’ is grammatical, even though the first object goes with the first verb and the second object with the second verb (a so-called cross-serial dependency).
We know this because the first verb insists on a dative object and the second verb an accusative object; the sentence cannot mean ‘Jan says that we helped the house paint Hans’, as its analogue in Dutch (which permits cross-serial dependency but lacks case-flagging on nouns) conceivably could. Indeed, this very sentence is the prima facie evidence that human language cannot in general be described by a Chomsky Type 2 (context-free) grammar.
Not that I know any Dutch, but I reconstruct the sentence in question as Jan zegt dat we Hans het huis hielp schilderen. GT knows how to cope with the cross-serial dependency, though it does not, going in the reverse direction, render the English sentence like this. Instead we get the properly tree-shaped Jan vertelt dat we Hans hebben geholpen met het schilderen van het huis ‘Jan said that we helped Hans with the painting of the house’. I’m not sure why GT prefers the present perfect to the preterite here, as it is definitely still in use in Dutch.
ObEty: Schilderen started out as ‘paint a shield’, and was later extended to other kinds of painting.
No, help.PRES-PL; very much like Old English, mainstream Swiss German has a “unity plural” in /d/.
More tomorrow, but so far it seems to me like the example is merely a tree with two crossed lines. That’s still a tree, very much unlike the loops in the example I posted.
I’m not sure why GT prefers the present perfect to the preterite here, as it is definitely still in use in Dutch.
In my experience, Dutch uses the present perfect more often than English, but less often than German. This may simply be one of the cases where it goes with German, where you wouldn’t use the simple past in this case if it’s meant as a simple report of your activities.The simple past would only be used if the sentence were meant as background information for some other past event (like “what were you doing when the barn burnt down?”). I am talking about literary German here, or a variety of Northern colloquial German that’s relatively close to the literary Standard; Southern German dialects don’t use the simple past at all, and Southern colloquial Standard German very rarely.
@jc
The dative in em Hans does NOT come from the verb helfen. This is a common (South) German construction, because genitive is moribund. So wir haben dem Hans sein Haus anzustreichen geholfen. Here helfen is intransitive. DM may explain this better or even that I am wrong☺.
@PlasticPaddy, you’re wrong. Consider this: Der Jan säit dass mer em Hans em Ruedi si huus hälfed aastriche.There you have both kinds of dative, help-dative and possesive-dative.
@PP
If it was the possessive dative it would be ’em Hans sys huus’, sys being the neuter singular nominative/accusative form of the 3rd person masculine possessive adjective. es is the indefinite article (not, as JC’s translation has it the definite article – that would be s in Zurich dialect, which this appears to be.
Western Swiss German dialects, such as Bernese, have the same distinction as standard German in the plural of verbs – the same in 1st and 3rd person, with 2nd person different (-e, -et, -e). The Upper Valais has different forms for all three (-e, -et, -ent), while a small area in Unterwalden has the same form for 2nd and 3rd person, with a different one for the 1st person (-i, -id, -id).
‘Jan’ as a name in a Swiss German sentence feels wrong, even with the necessary definite article added.
JC, what is basement on the west side of the house is ground level on the east side
I love that. There are lots of houses like that in San Francisco on the sides of hills; you enter a little cottage from the street, and by the time you look up from the back garden it’s a towering mansion of four storeys. Vice versa on the other side of the street, of course.
I hate basements. If you want more space, build an extra floor above ground.
There’s a good reason to have a basement (a.k.a. “a storm cellar”) if you live in the approximately 1/2 of the continental US where tornadoes are frequent phenomena. LH, I take it, lives in a fairly low risk zone, but if his house was built more than 30 years ago, a basement still would have been absolutely standard, not some kind of add-on. “No basement” in the US usually denotes 1) west of the Rockies or 2) very recent vintage.
I used to work in a building which had four floors, all of which were above ground, but the lowest floor was still counted as a basement, because the main access was on the next floor up via a walkway along the front (there were doors at ground level, but they were fire escapes, though we did have to use one as an entrance for a time while a lift was being installed). It kind of made sense, particularly as the ‘basement’ did have a dark basement-y feel to it by comparison to the rest of the building – it had plenty of windows, but the walkway at the front and a steep slope at the back cut out a lot of the light.
Just like ancient Germans applied the word for ‘bowl’ to ‘head’ when they met people with heads, or like New Guineans likened hair to grass, when they met people with hair.
In Stockholm which is essentially built on uplifted skerries, it is not unusual to enter an office building through the reception and finding yourself on the fourth floor (by elevator numbering). The goods delivery platform may well be on level -1 on the other side of the house.
(For similar reasons having split level crossings between major streets is not unusual and you will often find escalators or outdoor elevators for pedestrian use. Learning to get from here to there by car takes a long time, even when ‘there’ is just 8 meters up or down).
@k.ph
To be correct, it was the degenerate late Latins who, on forgetting their civilised forebears had once met people with heads, applied the word for jar to the heads of the detested barbarian invaders. These then discarded their word for head in favour of a corresponding word for cup in their own barbaric language and the mistaken belief that this was how civilised people spoke!
Isn’t there are word for that type of half-basement floor? It escapes me at the moment. I’m not sure if I find those an improvement over an actual basement, though. By the way, this discussion reminds me of a relative, who used to go to the “basement” to get things, the basement being not another floor but an extra pantry outside of the kitchen.
I’m intrigued by those people who used to not have heads, or hair. At least they had bowls.
Isn’t there are word for that type of half-basement floor? It escapes me at the moment.
Cryptoporticus
According to Wiki and M-W, that’s an ancient Roman thing.
I designed one and I’m not an ancient Roman.
How about grotto. Frederick the Great has or had a grotto at Potsdam.
Moa: Isn’t there are word for that type of half-basement floor? It escapes me at the moment.
sokkeletasje
Built on every other Norwegian house, and certainly if it’s a catalogue house, since the … late sixties, I think, and well into the nineties. Also on flat ground.
sokkeletasje
Googling that gets me the German equivalent Sockelgeschoss, and googling that gets me this discussion, which suggests “semi-basement.”
Yeah, they’re all over the place. They have tiny windows too high off the ground and concrete, carpeted floors that smell of cat pee. I’d rather have a grotto or a cryptoporticus, myself.
@Moa: The term in American architecture is “daylight basement.” Normally, this refers to the situation where the level is underground at the front of the house but above ground at the rear (so the opposite of what John Cowan describes).
When I started this blog I had no idea I was going to learn so much about architecture.
Talking of which, there are far too few Frenchmen in Space in science fiction.
In the TV show Dark Matter (which is both very good and very Canadian; one episode has the wait-time in a walk-in medical clinic as a plot point) has Quebec French in Space. The actor in question is one of the few Anglo-Canadians to have a steady career in French-language TV too.
There’s Riverworld, of course; every Frenchman ever born is there, including Cyrano de Bergerac.
They have tiny windows too high off the ground and concrete, carpeted floors that smell of cat pee.
I once lived in such a place (minus the cat pee aroma, thankfully) on Capitol Hill in DC. They were always described in real estate lingo as “English basements.” I can’t imagine why.
“English basements.” I can’t imagine why.
Damp? (I can say that since we’re both English emigrés.) Do they have French windows?
The term in American architecture is “daylight basement.”
American construction or American real estate, perhaps. It’s not an architectural term.
Casemate is a good word for the windows in one of these spaces though I’ve never heard it used for modern arch. I built a house in Hamburg with kasematten, though. They had a lap pool in their basement (it was the ’90s).
Nowadays every Russian family in London has a newly dug basement in their Victorian house. Depending on how many levels they can afford it has a pool, a cinema and some parking spaces – that’s really all you can do with public space with no natural light. Some have servants quarters (servants don’t need windows). They could I suppose have a library but I’ve never seen one. Much of the Royal Borough of Kensington & Chelsea has been undermined by all this digging and there’s discussion about putting an end to it (the basements not the Borough, although…)
No French windows. Often they had large bay windows at the front. A typical English basement is the bottom story of a late 19th century row house, often 4 stories in all, with an entrance that was half a staircase (I don’t know if that’s a bona fide architectural term) down from street level. The main entrance to the house proper was half a staircase up. The stairs are cast iron, as a rule.
One of my brothers once lived in a similar arrangement near Regents Park, in the bottom part of one of those splendid Georgian (?maybe) houses, where either the servants used to live or the coal used to be delivered. (Didn’t the servants usually live in tiny attic rooms?). And they also exist in Hove, in those large terraces just off the sea — the ones built with bungaroosh.
Shebs was an error for Shieber; evidently I conflated Stuart Shieber with Stanley Shebs, as I’ve just read a paper by each of them.
In any case, the gloss is as printed by Shieber (except that I moved the case flag from the nouns to the articles) so I accept neither credit nor blame for it. I can say that he tried out this and 61 other sentences on his four informants, all of them speakers of Zürich dialect (though one claimed to have some contamination from Bernese). Some were grammatical, some ungrammatical, but nobody seems to have complained about a missing article for Jan (though perhaps they were not given the introductory part of the sentence).
The paper’s only 11 pages, people might as well read it.
I’m not an ancient Roman
True. Although if the ancient Romans had thought of putting rebar in their concrete, the matter might be in doubt. The Château de Vincennes has cast-iron rebar, still unrusted after five hundred years.
I’ve now read the paper. It says Swiss German (and, by citation, Dutch and apparently Standard German) have a property called weak non-context-freeness. I don’t know what that is, but it doesn’t seem to mean “unable to be diagrammed as a tree”; it is mentioned in passing that all these languages are “linear-parsable”, which to me sounds a lot like “treelike”.
The paper does contain this priceless paragraph (in section 4.4, after performing a reductio ad absurdum):
On, then, to the details.
Does anybody know why the case-marking is so important to the argument? Here are the two ways to render the sentence in Standard German:
Jan sagt, dass wir Hans helfen, das Haus anzustreichen.
Jan sagt, dass wir Hans das Haus an(zu)streichen helfen.
(…Three ways. Among the ways that…)
In both (the first would generally be preferred), Hans is “marked” as dative by the word order. He paints the house, we help him; even without articles or endings there is no other way to parse this.
More fun with word order:
dass wir Hans helfen
dass wir Hans anstreichen helfen
dass wir Hans das Haus anstreichen helfen
dass wir die Kinder Hans das Haus anstreichen helfen lassen
that we.NOM the.NOM/ACC children.NOM/GEN/ACC Hans.NOM/?GEN/DAT/ACC the.NOM/ACC house.NOM/DAT/ACC paint.PRES.INF/1PL/3PL help.PRES.INF/1PL/3PL let.PRES.INF/1PL/3PL
Again there’s no other way to parse it: Hans paints the house, the children help him, and we let them, so the children are in the accusative, Hans is in the dative, and the house is accusative again.
Unfortunately, that’s the translation in the paper, and so is the past-tense “helped”.
I’m not an ancient Roman.
In interreti, te canem esse scit nemo.
dass wir Hans anstreichen helfen
Again there’s no other way to parse it
There is doch another way to parse it: when Hans is the children’s name for their wooden horse.
“Parsing” is merely a technique for finding sense in a sequence of words. Even when that sense is nonsense (“the furious sleep of reason produces monsters”).
Throwing pearls before swine is a good way to discover whether any of them has the potential to become a good investment banker.
Hans das Haus
Bernd das Brot
Wir helfen Hans anstreichen und Bernd das Brot bestreichen.
The Château de Vincennes has cast-iron rebar, still unrusted after five hundred years.
Thanks for that, John. A recent discovery. I found more here (à la p. 10). Iron & lead and unrusted because perhaps only the lead was exposed to corrosion. It’s not quite the same as rebar because a) no concrete and b) the masonry & metal are side-by-side but not quite working in tandem as they would in reinforced concrete. Apparently it was very expensive and perhaps done more to save time while the mortar dried than to strengthen the masonry. But they really knew what they were doing (they used chains at the top, we did catenaries the other day) and there’s an important French tradition here. In 1780-something Soufflot drew similar bits of cast iron next to the stone reinforcing the entablature of Ste Geneviève (soon to be confusingly renamed the Panthéon) – you can blow both those images up and toggle between them (quite fascinating for some of us). Structural analysis in a recorded form started about that time (see fig.1 of this drawing of St Peter’s from 1742 ) and the Germans subsequently got involved there, but reinforced concrete was developed most of all by the French. Auguste Perret’s use of Hennebique’s reinforcement system at the Rue Franklin Apts is the standard example (and then Le Corbusier worked for Perret – and ignore his being Swiss and Perret Belgian because we’re not talking passports but culture).
If the iron and lead are in contact, I believe only the lead will corrode because electromotive table.
I love that. There are lots of houses like that in San Francisco on the sides of hills; you enter a little cottage from the street, and by the time you look up from the back garden it’s a towering mansion of four storeys.
Since Edinburgh’s Old Town is pretty much built on the side of a cliff, you can go in the ground floor of the City Chambers (for example), walk through to the other side, and find yourself on the seventh floor.
Does anybody know why the case-marking is so important to the argument?
Per one of Pullum’s essays, it’s because the Dutch sentence could be parsed “Jan said that (Hans (the house helped) paint)”, which is properly hierarchical even though the semantics are absurd (modulo Hans the hobby-horse). Thus Dutch is not (strongly) context-free. But the Swiss German flags make even that stupid parse impossible, unless you want to believe that flagging is itself semantic and doesn’t affect parsing; Shieber considers and rejects this view in section 4.3. This is why Swiss German is said to be not (even) weakly context-free. The distinction between strong and weak is non-formal; it has to do with whether the hierarchical parse exists but is meaningless, or it doesn’t exist at all.
Linearly parseable just means that you never have to have an unbounded amount of lookahead to parse the grammar correctly.
Pullum’s view is that despite exceptions like this one, or the problem with large numbers (the zillions must be preceded by the zillion zillions, which must be preceded by the zillion zillion zillions, …, where zillion is the largest monomorphemic number), or the Mohawk incorporation problem (you can start with a Mohawk verb that incorporates its subject, nominalize it, incorporate it into another verb, … — but then an explicit subject must agree with the innermost subject, indefinitely far away), there is so much of natural language that is context-free that the rest can be safely ignored.
San Francisco
The hill in which my house is set is in fact the dam which makes our pond a pond instead of a swamp: it’s maybe 10 feet above the level of the road.
Thanks, JC. You acquire Merit by linking to papers by Geoffrey Pullum, and on our own level we acquire Merit by reading them.
I always enjoy his takedowns of Chomsky’s (egregious) mathematical errors particularly.
(And I liked “Elster’s argument depends on a confusion between grammar and arithmetic”, too.)
“no other way to parse it”
How about “Han’s children”?
Ha! Clever! 🙂
Alas, even inserting commas is no match for a good blaster by your side:
dass wir, die Kinder Hans, das Haus anstreichen helfen lassen
“that we, Han’s children, let OBJECT HIDDEN BY THE DARK SIDE help paint the house”
I’m not familiar with Star Wars and don’t know whether Han has any children. The only houses I remember are a desert home (unpainted), a bar (?), and a city in the sky (probably painted). I don’t know if the star destroyer’s outer surface would require anti-corrosive paint, not being exposed to much humidity or oxygen. Perhaps the inside would benefit from pastel colours to lighten the atmosphere.
(Dative object, that is.)
I was once a founding member of Foreningen til Hans eget bedste (the Association for Hans’ own good). As opposed to Foreningen til hans eget bedste (… his …), of course. In writing the capital letter of the personal name gives the game away, but spoken the difference is only a stød. (And by default stress, but even when giving hans contrastive stress it does not acquire a stød).