"Darmok"

[This is part of an ongoing occasional series about linguistics in science fiction.  Fair warning: spoilers.]  I've been meaning to write about the Star Trek: The Next Generation episode "Darmok" for years.  In fact, I had it in mind when I started this blog, and I promised that a post about it was imminent in September of 2004.  (Doh!)  Well, during our ongoing sojourn in Germany, we discovered to our horror that we'd run out of English-language TV.  Fortunately, I remembered that I've had an AVI of "Darmok" sitting in my hard drive since July of 2004, so we fired it up and watched it.

"Darkmok" is easily the most linguistic of the Star Trek episodes I've seen.  Unlike a lot of the SF I discuss, it's pretty well-known, so I'm a little nervous about writing about it—I'm sure you can find many, many essays about it online, but hopefully I have something to add.  For a comprehensive review, check out this page by Raphael Carter, which includes both a summary and a glossary of all the alien utterances in the episode (and from which I got the official spellings of most of the proper names).  I'm going to summarize the linguistic aspects of the story, then analyze where I think they're wrong—which is, unfortunately, pretty often.

Summary

A ship from an alien species called the Children of Tama has been waiting near the planet El-Adrel for three weeks, transmitting a signal containing a "standard mathematical progression" toward Federation space.  It seems they want to make contact, and the Enterprise has been sent to do so.  This may not be straightforward, though: seven times in the previous 100 years, Federation ships encountered Tamarian ships, but, as Data puts it, "formal relations were not established because communication was not possible."

When they get to the planet and open up a channel, this is what the Tamarian captain says:

Rai and Jiri at Lungha.  Rai of Lowani.  Lowani under two moons. Jiri of Ubaya.  Ubaya of crossroads, at Lungha.  Lungha, her sky gray.

Huh?  As Data observes, "the Tamarian seems to be stating the proper names of individuals and locations," but what's it all about? Picard replies with some polite suggestions about a mutual non-aggression pact but meets with incomprehension.  It's clear both sides think they're expressing themselves clearly, but no message is getting through.

The Tamarian captain and first officer argue, apparently over how to proceed.  Finally, the captain seems to come to a decision.  He pulls a knife from a sheath on his chest and another from his first officer's, turns toward Picard, holds out the knives, and intones, "Darmok and Jalad at Tanagra."  Before the Enterprise can stop them, the Tamarians have beamed both captains down to the planet.  Uh-oh, it looks like "Arena" all over again!  (Dude, make gunpowder!)

Down on the planet, Picard and the Tamarian eye each other warily. The Tamarian keeps repeating "Darmok and Jalad", holding out both knives.  Picard says he won't fight, and even when the Tamarian throws the second knife at his feet, Picard throws it back.  The Tamarian, seeming disappointed, says, "Shaka, when the walls fell."

The Enterprise can't beam up Picard because the Tamarian ship is interfering, and, predictably, it doesn't help trying to talk to them about it.  As attempts at retrieving Picard continue, Data and Troi try to work out a way to open up communications.  While studying the recording of the Tamarian officers arguing, it occurs to them to try searching through "all linguistic databases for this sector" for the proper name Darmok.  There are 47 hits.  But when they search for Tanagra, they notice that one of the hits is from same planet as one of the hits for Darmok.  (Apparently they don't have AND in their search engine.)  Data and Troi later summarize their research:

Data:	The Tamarian ego structure does not seem to allow what we normally think of as self-identity.  Their ability to abstract is highly unusual.  They seem to communicate through narrative imagery, a reference to the individuals and places which occur in their mytho-historical accounts.
	...
Troi:	Imagery is everything to the Tamarians. It embodies their emotional states, their very thought processes. It's how they communicate and it's how they think.
Riker:	If we know how they think, shouldn't we be able to get something across to them?
Data:	No, sir.  The situation is analogous to understanding the grammar of a language but none of the vocabulary.

Meanwhile, down on the planet, Picard is trying to figure out the same mystery.  When the Tamarian leaves his camp, Picard sneaks in to look for clues that might help him understand what the alien is saying.  The Tamarian comes rushing back, shouting about Darmok, and tries to press a knife into Picard's hand again.  Picard starts to refuse until he hears a scary snarling noise from behind the Tamarian. Looks like they've got company, and the Tamarian wants to fight together against the common foe.

In fact, that turns out to have been his plan all along.  Picard finally realizes that the Tamarians communicate by allusions to their mythology (although he calls this "metaphor", which I don't think is quite right).  Knowing this, he begins to make a little headway with the "vocabulary", discovering that "Shaka, when the walls fell" refers to failure, while "Temba, his arms wide" refers to giving.  The Tamarian wants to re-enact the story of Darmok and Jalad, who met on the island of Tanagra, fought together against the Beast of Tanagra, and left as friends.

His plan doesn't quite turn out as expected, and the creature on El-Adrel mortally wounds the Tamarian captain.  Before he dies, he tells Picard the story of Darmok and Jalad, and Picard responds with the story of Gilgamesh and Enkidu.  When Picard returns to the Enterprise to find the two ships fighting, he is able to defuse the situation by using his newfound understanding to speak with the Tamarian first officer, who acknowledges his account of the fight on the planet's surface by saying, "Picard and Dathon at El-Adrel."  Picard's not sure they've made new friends, "but at least they're not new enemies."

Analysis

There's something odd about the beginning of this episode.  It makes sense that an unfamiliar language would be, well, unfamiliar, so why are the crew of the Enterprise surprised not to be able to communicate immediately?  It turns out that in the Star Trek universe, nobody needs to worry about learning alien languages because they have a device called the "Universal Translator".  It was introduced in the original series, implicitly the first time they were able to communicate with a previously unknown species, and explicitly in the episode "Metamorphosis" in this conversation between Zefram Cochrane, Kirk, and Spock:

Cochrane:	What's the theory behind this device?
Kirk:	There are certain universal ideas and concepts common to all intelligent life.  This device instantaneously compares the frequency of brainwave patterns, selects those ideas and concepts it recognizes, and then provides the necessary grammar.
Spock:	Then it simply translates its findings into English.
Cochrane:	You mean it speaks?
Kirk:	With a voice, or the approximation, of whatever the creature is that's sending it.  Not 100% efficient, of course, but nothing ever is.

This dialog takes place while Spock is tinkering together a Universal Translator with a screwdriver (which he can do because Spock is The Man).  It's a metallic cylinder about the size and shape of a lightsaber.  Even though it's hand-built and impressively compact, the resulting device is so sensitive it's able to detect and reveal that the amorphous alien energy creature they use it on is female—and because this is Star Trek, that's an imporant plot point.  (At least Kirk doesn't end up putting the moves on her.)

The Universal Translator is pure Treknobabble.  It becomes clear over the course of the series that it works in both directions, which means that it's able to both read thoughts in and project thoughts into the neural architecture of any kind of creature, from bumpy-foreheaded humanoids to life-Jim-but-not-as-we-know-it. That would imply the Star Trek universe has working mechanical telepathy, but nobody ever seems to mention that¹.  This is because the Universal Translator is nothing more than a fictional convenience.  It exists to save the producers of a weekly TV show the trouble of making up a new language for every new species, and as such it doesn't bear much thinking about—unfortunately, though, that's just what "Darmok" forces us to do.

Note that the Universal Translator clearly works, at least partly, on the Children of Tama.  When they speak, the crew of the Enterprise don't hear an untranslated stream of Tamarese—"Darmok Jalad-qo Tanagra-nga".  Instead, they hear proper names (appropriately) untranslated, but all the function words, common nouns, and verbs are translated—"Darmok and Jalad at Tanagra", "Shaka, when the walls fell", "Temba, his arms wide".

So the problem is not that Tamarese is beyond the capabilities of the Universal Translator, which must mean that the Children of Tama's brains are running on more-or-less the same wavelength as everyone else's.  The problem, rather, is that they express themselves entirely using allusions to their mythology.  It's easy to imagine why this would be a problem for a person learning the language, but I don't see why it should bother the Universal Translator.  Data's speech above about the radically alien Tamarian self-concept is an attempt to finesse this point, I think—the Children of Tama think different—but I'm not convinced.  For guys who are supposed to lack self-identity, the Tamarian captain and first officer sure seem to argue like ordinary self-identifying people.  What's more, the UT evidently does partly work, which means it's successfully reading the Tamarian's minds—so when the captain thinks, "Failure!", but says (in Tamarese), "Shaka, when the walls fell", why would it produce a gloss of the spoken sentence instead of, you know, a translation?

The linguistic problems with "Darmok", however, run deeper than contradicting previous statements about the (admittedly magical) Universal Translator.  Let's assume the facts of the language situation are as the writers would have it—the UT isn't able to handle Tamarese because they speak in literary allusions.  Why would that prevent communication?  In spite of having starships for over a century in a galaxy teeming with intelligent species—at one point in the episode, Data mentions having personally encountered 1,754 in his Starfleet career—have the Children of Tama never figured out that other people don't share their mythology?  That seems like the sort of thing they'd have figured out from first principles before meeting anyone.

What's more, I don't believe in Tamarese as described.  With regards to language change, it must be in an extraordinarily unstable state.  Let's consider the problem of young Tamarians—the children of the Children of Tama, if you will—who, if the language works the way we're told it does, are presented with a seemingly insurmountable language-learning problem.

Think of Tamarese as two languages.  Tamarese-A is an ordinary human(oid) language with words and grammar and so forth, which, judging by the internal structure of the allusions we hear, has strategies for expressing common nouns ("walls", "arms"), coordination, possession, and other garden-variety linguistic phenomena.  Tamarese-B, their language of allusions, is encoded in Tamarese-A.  Are the Children of Tama consciously aware of both these levels?  If they are, they presumably know to speak Tamarese-A with children, at least until they learn enough of the canon of Tamarian mythology to get by in Tamarese-B.  But if that's the case, the whole episode falls down—after one or two attempts with Tamarese-B, wouldn't the captain give Tamarese-A a try?  Worse, when Picard says things in English to the Tamarian captain, the Universal Translator ought to be translating them into Tamarese-A.

The Children of Tama, therefore, must somehow use Tamarese-B exclusively, without realizing it's encoded in Tamarese-A.  If that's the case, we should expect Tamarese-B to last exactly one generation. When the next generation of children acquires it, they're going to have no idea what any of the allusions refer to.  Instead, just as Picard does, they'll simply learn it as a language with almost no grammar and a few very long words—for them, the Tamarese utterance corresponding to "Shaka, when the walls fell" will be a single lexical item meaning 'failure'.  Presumably, they'll then do what human children do in such a circumstance and invent a creole.  Tamarese-B, I think, really ought to have disappeared years before, but since failed attempts at communication go back a century, it must, improbably, have survived.

I also don't believe that a language based entirely on allusion is at all suitable for situations requiring precision.  If all you're doing is arguing about a general course of action, Tamarese-B might be enough, but how can you run a starship using it?  When the captain wants to tell the helmsman to go to warp factor five, does he say, "Darmok...uh...that time he went warp factor five"?  At the end of the episode, in fact, the first officer orders his ship to warp out of orbit with "Mirab, with sails unfurled", which is used several times in the episode to mean something like 'go'.  Shouldn't the helmsman reply, "Mirab-with-sails-unfurled factor what, sir?".

Come to think of it, what would you do if you wanted to use Tamarese-B to tell someone that a wall had fallen down?

"Shaka, when the walls fell"
[expression of sympathy at your failure]
"No, listen—Shaka, when the walls fell"
[blank look]
"Dude, seriously—Shaka, when the literal walls literally fell"

I imagine that would get old fast.

Another weakness in the episode is that the only reason the crew of the Enterprise are able to start puzzling out what the Children of Tama are saying is that, as near-human bumpyheads, they have familiar human gestures, facial expressions, and tones of voice.  For example, Picard is only able to guess that "Shaka, when the walls fell" means 'failure' because the Tamarian captain hangs his head and looks frustrated when he says it.  The whole episode would have gone very differently, I think, if the Children of Tama had looked like hortas.  There's also the problematic scene in which Troi and Data are able to look key proper names (Darmok and Tanagra) in some kind of pan-Galactic linguistics/mythology database.  If they can do that, the Children of Tama have already managed to communicate with someone, either to learn that myth or to transmit it.

Even when I originally watched it, I found "Darmok" frustratingly poorly thought out, and knowing more about language has only made matters worse.  This is partly because I find the core story pretty appealing:

• Two peoples meet.
• They encounter a communication gulf that both want to bridge.
• One side acts to do so, but in a way that is easy to interpret as an attack.
• Luckily, through persistence and sacrifice, good will overcomes paranoia and communication is established.

That sounds like it would make a nice, optimistic episode of Star Trek, doesn't it?  Unfortunately, in fleshing out this skeleton, the writers constructed a linguistic situation that's simply unbelievable.  It's not that a language made up entirely out of allusions is unworthy of fictional exploration.  Raphael Carter suggests Tamarese is similar to the language of the Ascians in Gene Wolfe's Book of the New Sun, and I wonder if either of these might have been inspired by the four-character idioms that famously give students of Chinese so much trouble.  But stories about such a language-of-allusions just don't fit into the Star Trek universe because it can't be squared with the Universal Translator.  It occurs to me that a lot of these problems could have been avoided if the Children of Tama, instead of being aliens, had been a lost colony of humans—that would at least have explained why their speech was partly comprehensible and why body-language cues would help.

Still, it's not a bad bit of sci-fi TV.  That's due in large part to Patrick Stewart, who seems to have the mutant ability to take any script, even one which doesn't entirely make sense, and infuse it with sincerity and meaning.  ("That's why they call it acting.")  This helps some, but for me, it's not enough to save "Darmok" from its flaws.

[The Wikipedia page for "Darmok" has some good links at the bottom, including a page about it on Memory Alpha, the offical page on startrek.com, and a non-canonical work of fictional non-fiction by Jean-Luc Picard about the incident.]

¹ Recall, however, that in a couple of original series episodes, Federation courtrooms are shown to have a working lie detector.  This technology, too, would require a particularly thorough and reliable form of mechanical telepathy.

Comments

[JS Bangs]

Fascinating discussion of a STNG episode that I've never seen. Since I haven't seen the episode, I don't have much to add, except to say that it reminds me of the story by U.K. LeGuin called "The Nna Mmoy Language". If you haven't read the story, the premise is that the Nna Mmoy people speak a language in which "words" are only bundles of connotation, and are almost infinitely malleable by context. The language is impossible for the "transomat" to handle, and very difficult for adult humans to learn. However, LeGuin says that children learning to speak the language speak much more "normally", and only acquire the adults' connotative allusiveness in adolescence.

This is basically the same as your Tamarese-A and Tamarese-B scheme, but I don't find it inherently implausible for the Tamarese or the Nna Mmoy. Obviously the adult language has enough structure for the children to learn a basic syntax and vocabulary, and once that's learned the children can bootstrap themselves into the adult language. But this has all of the problems you mentioned above. Let's assume that the Tamarese consider it insulting to speak Tamarese-A with another adults, so much that they never attempted it in the past--that solves one side of the problem. But presumably the UT can create working Tamarese-A, and the transomat can create infantile Nna Mmoy, so at least communication in one direction is possible.

That gives me an idea for a better story: communication seems to be successful between the groups, but the Tamarese are dismissive and condescending because the humans are all using baby-talk.

[Lance]

Oh, thank heavens, I thought it was just me. I read Raphael Carter's glowing approval of the episode well before I saw it, and then when I did see it I thought, "Uh, what?"

You hit on pretty much all of the objections I thought of (Starfleet knowing their myths; using allusion to say "Hey, I think that screw over there needs tightening"; the language acquisition problem), and many others, thereby saving me the need to write this up myself with less thoroughness and less eloquence.

So: excellent post, and thank you.

[Doug Sundseth]

I point that is sort of implicit in your analysis but not really explicitly discussed is how one would write the story of "Shaka", without a Tamarese-A. I suppose it might be written with further allusions to other literary constructs, but I can't figure out how this could be completely recursive. I think such an allusory language must have some sort of implicit or explicit allusion database and that database must include some sort of literal primitives.

All of which is to say, "I agree." (Another stereotypical translator gag, of course.)

8-)

[Hamilton Lovecraft]

stories about such a language-of-allusions just don't fit into the Star Trek universe because it can't be squared with the Universal Translator

Or at least, with the explanation we're given for how the UT works in "Metamorphosis".

Besides mechanical telepathy, the UT also needs to incorporate a fairly sophisticated artificial intelligence to do its job -- in "Metamorphosis" in particular it has to synthesize human/English inflections, speech patterns, etc. from a non-verbal communication pattern.

I can buy that the AI might decide to stop its decoding at the "Shaka, when the walls fell" level -- assume it's trying to preserve some sense of the flavor of the communication, not merely the most thorough and dry translation. What doesn't make sense is that the Tamarese couldn't figure out how to use the vocabulary used for phrases like "when the walls fell" to communicate more effectively.

[Steve]

Here's an interesting line from another blog:

"Perhaps more lawyers need to see this episode, for the Tamarian language is exactly what we sound like to laymen. We communicate with each other by shorthand references to complex legal doctrines that are, without a fairly full explanation of context, misleading when reduced to a short summary."

[Anatol Stefanowitsch]

As much as I like your incisive analysis (and the analyses of others before you) of the shortcomings of Darmok from a linguistic perspective, I don’t think that it does justice to one of my favorite Star Trek episodes. It is one of the few episodes (and one of the few first-contact stories in general) that attempts to draw people’s attention to the language barrier that must be overcome in any first contact situation (human-human or human-alien). This problem may be obvious to us, but it is not at all obvious to non-linguists, as evidenced by the fact that people generally do not find anything strange about first-contact stories that ignore the problem of language completely (and there are plenty such stories, most of which do not even bother to invent a ‘universal translator’).

In order to be able to focus on the linguistic problems posed by a first-contact situation within the logic of Star Trek universe, the authors of this episode had to find a way of removing the universal translator from the equation, so they invented a language for which the universal translator would be useless. This language turns out to have interesting properties, because in order to derail the universal translator, it must violate the folk theory of language on which this device is (partially) based. Like most current machine translation systems, the universal translator assumes that (i) words correspond directly to meanings (lexical concepts or whatever) and that (ii) utterances are created by applying grammatical rules to words. The universal translator applies this folk theory to the Tamarin language and fails. In doing so, it brings to attention an aspect of language that non-linguists are generally unaware of: non-compositionality.

Human languages generally provide speakers with the resources necessary to communicate ideas in a fully compositional way. Yet speakers choose to use all kinds of non-compositional structures whose full interpretation often requires in-depth knowledge of the matrix culture (including, but not limited to, proverbs, idioms of various degrees of flexibility and abstraction, quotations, ‘snowclones’ etc.). As anyone who has ever looked at authentic data knows, such non-compositional structures are at least as frequent as compositional ones. Tamarin simply takes the reliance on non-compostional structures to an extreme, but it is possible to find stretches of English conversation that do not sound that different from Tamarin. By presenting an extreme example of a fully non-compositional language, Darmok allows non-linguists to think about this facet of communication (in fact, I sometimes use it in introductory linguistics classes to introduce the idea of non-compositionality).

From a linguist's perspective, of course, I prefer descriptions of alien communication systems that remain truly alien (as, for example, in C.J. Cherryh’s Forty-Thousand in Gehenna) or descriptions of first-contact situations that pay proper respect to the linguistic expertise and the field work techniques that would be necessary to establish a basis for communication (as, for example, in Mary Doria Russell’s The Sparrow). But of course most science fiction (and fiction in general) is not made for linguists. Ok, so as linguists we know that Tamarin could never be acquired, that it cannot be used to coordinate the operation of a space ship, that the universal translator is an impossible device, etc. But if we switch off this knowledge for an hour, Darmok presents an interesting linguistic situation (and some emotionally charged moments for those of us who admire Captain Jean-Luc Picard. I mean, “Picard and Dathon at El-Adrel” -- how much of a stickler for linguistic plausibility do you have to be in order not to be moved by that?).

[Glyn]

I seem to remember an early story in the General Hospital series by the Northern Irish writer James White, about some human who started a war when 1st Contact went drastically wrong - not having a universal translaror he used gestures that were extremely offensive to the aliens. Can anyone remind me about the specifics of this?

[Ran]

I agree with Anatol Stefaniwitsch.

In general, the Star Trek franchise has very consistently treated some concepts as not requiring consistency. For example, look at all the different versions of time travel that appear in the various series (half of them in Voyager alone); it's a matter of debate how time travel paradoxes would be resolved in a Universe that actually allowed them, but I think it's pretty obvious that no real Universe could allow all the different ways they're resolved in Star Trek (causality loops, anticausality loops, the confusing causality/anticausality blend in the early episode of Voyager where they find a planet that's been destroyed by a polaron explosion, and so on).

Incidentally, with the Universal Translator specifically, the other series negate the telepathy approach: there are various episodes of various series where the translation matrix is applied to recordings, written text, and so on; there's one episode of DS9 where the Universal Translator needs to gather more data, so they have to keep the aliens talking for a while until the Universal Translator is of help; and many episodes of Enterprise involve Hoshi's wranglings with the Universal Translator and the linguistic database (both of which she developed).

[jds]

Well, I think those who try to think of this episode from the perspective of linguistics are largely missing the theoretical point that it makes. I'm not at all a fan of Star Trek, and it was largely by accident that I saw it in the early 90s while an undergraduate in literature. I immediately recognized its origin, as did the roomful of lit-types watching it with me. You'll need to read a short but brilliant little parable by Nietzsche whose titled is translated something like "on truth and lie in an extra-moral sense". It's a little myth about the origin of language: that at first the meaning of words is purely metaphoric -- everything is a metaphor -- but over time and through usage these metaphors become literalized, "worn down", and their original arbitrariness is forgotten -- and then the forgetting is itself forgotten. We referred to this kind of theory as "rhetorization of grammar". So, like the very best science fiction, this episode acts out an impossible experiment: what if we were to go back in time and re-encounter the figural nature of all our signs?

[Sally]

Greetings, her arms wide! I know you recognize you're late to this observation. But if I recall correctly, from the moment "Darmok" aired, this linguistic "defect" or lapse in credibility had been widely commented on and joked about, at least by word of mouth if not on-line. I remember it well, as I was an engrossed viewer then. And a science fiction writer who also wrote briefly for ST:TNG ("Hollow Pursuits," 1990; "Babel" 1992).

What "Darmok" did that was worth noting for me, despite its simplicities and the way action and explanation often move at a snail's pace on TNG, was to reference the story of Gilgamesh. It's linguistic experiment as I justified it then was not to be taken so darn literally, but seen in context with its message about the way credible alien peoples may have tried and failed to communicate (minus that annoying UT)--as you rightly point out. It's message for me was that one chieftain thought it worth risking his own life to make that connection with the "wild bald man" ;) and force him into a struggle with the Bull of Heaven.

Right, you can't say "pass the Arcturian sea salt" in Tamarian, or "Keptin! We are having trouble wis the power conwerters!" At least not at that level of development. Indeed, the on-line pages that pore over Tamarian for its meanings look as though its contributors think it's a believable "conlang." It's a symbolic one, that had a little structure to it. And it should be taken as such.

Boy I admire the ones who thunk it up and aired it. It transcended, for the moment, the earnest quality of verisimilitude that most of TNG aimed for.

Sally Caves

[Sally]

What Ran says! I guess I'm a lit-type, too, as well as an amateur linguist, so I enjoy these philsophical discussions. I very much like the Nietzsche parable. More recent linguists talk of image-schemas dictating the very way we conceive of the universe. So EVERYTHING, basically, is a metaphor, c.f. George Lakoff, Mark Johnson, etc.

Ran wrote above:
"You'll need to read a short but brilliant little parable by Nietzsche whose titled is translated something like "on truth and lie in an extra-moral sense". It's a little myth about the origin of language: that at first the meaning of words is purely metaphoric -- everything is a metaphor -- but over time and through usage these metaphors become literalized, "worn down", and their original arbitrariness is forgotten -- and then the forgetting is itself forgotten."

Right on. (metaphor: argument as cap on the bottle.):)

[Marcos]

JS Bangs:

"Nna Mmoy"? Seriously? Did LeGuin write this before or after the prominent association of one Lle Nnard Nna Mmoy with SF? :)

[jds]

Sally,

How I love the internet. Here's a link to the Nietzsche piece:

http://www.geocities.com/thenietzschechannel/tls.htm

I hadn't read it in years. It's wonderfully tonic, esp in contrast to all the Alec Marantz I'm having to work through these days, eh. (Not that I dislike DM, but it's a different style...)

jds

[David J. Peterson]

I'll resist the urge to comment on the Bickertonian creolization myth that's now accepted as fact by the "science" of linguistics, but I would like to question the impossible Universal Translator. Should there be a way to fool it? There are numerous Star Trek episodes where aliens speak their own language, and it comes through as their own language. Is there some cue that they can give off to let the UT know that "I'm speaking my own language for locutionary purposes?" (E.g., if a Klingon was on trial, and they wanted to establish, for the record, that he could speak Klingon, by having a conversation in Klingon with him.) What I always took from this episode is that the Tamarians, like Lugones's chimps, had a language, but refused to use it. In that refusal, couldn't there have been something to either trip up the UT, or to let it know not to translate the meaning, but the words? (The question of "why" remains, of course.)

A similar question could be asked about the episode where the Enterprise encounters a crew of "dumb" aliens. Here, the UT is apparently able to decipher the language the aliens are using, and tell that they aren't using it very well. That's a pretty clever machine.

Oh, and since gesture was brought up, an interesting question: might not the UT be able to interpret and encode gesture, so it made more sense to the viewer? If not, I'd bet it'd have a very tough time translating signed languages.

[Christopher Culver]

FWIW, the language of the Ascians is clearly based on the Chinese quoting from Mao's "little red book" as well as from Orwell's Newspeak. I doubt Wolfe knew about four-character idioms at the time.

[Lance]

The thing is, there are of course various ways of viewing anything, be it Mt. Fuji or an episode of Star Trek. From the perspective of the literature major, there may be a great deal to like in this episode; from the perspective of the Trekkie, there may be some nice points and some frustrations. From the point of view of at least some linguists--at least myself, and, I gather, TstT--this episode was a dismal and painful failure.

With respect to Mr. Peterson's observations: one of the real frustrations, for me, of the UT was its inconsistency. I'm willing to accept it (as a sf-tv viewer, though not as a linguist) as a bit of magical technology; but the fact that it happens to stop working when some Klingons are speaking, sometimes, kicks me just a little out of the created reality. But it is an interesting thought that perhaps Tamarese-A exists but, for sociological reasons, they refuse to use it for higher-level discussions. It'd have to be a pretty strong taboo, of course, and I think a lot of the other problems TstT highlights still exist.

The UT, incidentally, definitely cannot translate signed languages; see "Loud as a Whisper", in which Troi is forced to translate for a deaf negotiator. (Notable, too, for actually casting a deaf actor.)

(Also: whoa, the person who created Barclay stopped by to comment? How Frickin' Cool is that? Must remember to track down your other writing, Sally.)

[Sally]

Hi JDS; no way to go back and edit my posts! It's confusing when the name of the poster goes under the line of the prior post. I'll get used to this! Thanks for the link!

Sally

[Sally]

"whoa, the person who created Barclay stopped by to comment? How Frickin' Cool is that? Must remember to track down your other writing, Sally."

Hey there, Lance! Thanks! I was Barclay, daring to play in another man's sandbox. Got a little sandy.

Not much to see in paper print (a couple of stories in F&SF, Terra Incognita, etc.). Most of my inventive liquor has been pored into vast pools on-line. Or into essays required for scholarly publication in the halls of Higher Learning.

Sally Caves

[Sally]

PS Lance: I agree with all the comments above about the irritations of the UT. But these writers were churning things out at a dizzying speed to produce one of the most popular television series on air; they even joked about the UT themselves. Where you needed to explain something in technical terms, you'd write in the dialogue and they'd send it to their tech guys who would provide the appropriate babble in record time. All in the interests of getting a wrap in a week. The UT was the least important thing on their agenda. What they were thinking of was: originality of the plot and its message, how effectively it was written, A and B story (very important!), and getting everything into 43 minutes. Linguists they were definitely not. Skilled writers? It varied from episode to episode. "Measure of a Man" by Melinda Snodgrass was one of the stellar episodes they had: taut, controversial, moving, award-winning.

However the speculations that are being made above about the limitations of mechanical translation are fascinating, and so are the references to other more interesting language inventions, such as the ones Anatol lists above. I agree with Anatol, though, about the philosophical importance of this episode and its attempt to "trick" the UT into making communication difficult for first contact. I was intrigued, too, by his remarks about non-compositional language. "Darmok" was a class-act episode; the actor who played the alien had the most expressive face and conveyed a complex and sympathetic personality in so few words and gestures.

Anatol wrote:
"But if we switch off this knowledge for an hour, Darmok presents an interesting linguistic situation (and some emotionally charged moments for those of us who admire Captain Jean-Luc Picard. I mean, “Picard and Dathon at El-Adrel” -- how much of a stickler for linguistic plausibility do you have to be in order not to be moved by that?)."

Well exactly! S.

[The Ridger]

The thing I always loved about the UT was the way it knew to leave some words in the original language.

[basilikon]

Just a little bit of ST trivia here, but at the risk of sounding condescending (certainly not intentional here), I thought I would share this. Paul Winfield, the actor who played Dathan (the Tamarian captain), also played Captain Terrell (the CO of the USS Reliant, which was hijacked by Khan Singh in Star Trek II: The Wrath of Khan). I was surprised when I realised who was under all that makeup in 'Darmok'! Sadly, he passed away in March of 2004 from a heart attack. I found him to be a rather accomplished screen actor, even though my only exposure to him was in the world of Star Trek.

David

[The Tensor]

Wow, lots of comments. It's interesting how the first few seem generally to agree with what I've written, then there's a wave of more critical comments. I'll bet that represents first the linguistically-oriented regulars, then people following links from more science-fiction-oriented sites.

Anyway, I want to address some of the points raised:

• I haven't read LeGuin's "The Nna Mmoy Language", but I'll add it to the Linguistics in SF pile.
• I'm aware that there have been revisions over the years to how the Universal Translator is supposed to work. In fact, one of the reasons I'm continuously tempted to rent and watch Enterprise on DVD is that there's a character whose job it is to build a much more linguistically-plausible UT. But that doesn't change anything: either there should just be a magical UT, in which case I will obligingly suspend my disbelief, or else the way language is treated should be consistent and sensible. You can't expect me to not examine linguistically an episode whose central speculative concern is language.
• In writing this post, I considered the idea that the Children of Tama can communicate in Tamarese-A, but for some reason refuse to, but I don't think it fits with the episode. Dathon is clearly motivated to communicate, so much so that he risks and loses his life to achieve it. I can't believe that, if the answer was as simple as speaking baby-talk to the Enterprise, he wouldn't do it.
• I agree that the episode was an effective story—I say as much in the last few paragraphs of the post—but I'm writing about "Linguistics in SF" here, and the linguistics wasn't handled very well.
• I'm not sure what to make of jds's "I think those who try to think of this episode from the perspective of linguistics are largely missing the theoretical point that it makes". What other perspective is more appropriate for thinking about theoretical points about language? Maybe it's people who don't think about language from the perspective of linguistics who are missing something...
• Sally, your DS9 episode "Babel" is about language! Excellent! But: "More recent linguists talk of image-schemas dictating the very way we conceive of the universe. So EVERYTHING, basically, is a metaphor, c.f. George Lakoff, Mark Johnson, etc." Lakoff's ideas are quite controversial within linguistics. You might try Semantic Compositions' skeptical look at Don't Think of an Elephant (1 2 3 4 5).
• Re: Paul Winfield. Aha! Both times I saw this episode, back when it originally aired and again last week, I spent the whole time thinking, "Hmm, that guy seems familiar..."

[JS Bangs]

LeGuin's story is in her recent anthology _Changing Planes_, if you're going to read it. All of the stories are good, so it's worthwhile. My copy is signed by LeGuin herself :).

Sally: the Sally Caves of Teonaht fame? Why, I know you from CONLANG!

[Lance]

In fact, one of the reasons I'm continuously tempted to rent and watch Enterprise on DVD is that there's a character whose job it is to build a much more linguistically-plausible UT.

Spare yourself. I kind of liked the show in spite of myself, what I saw of it; but linguistically? There's an episode where Hoshi declares, "The grammar sounds bimodal"--which is to say, having a linguist on-board just means that the technobabble extends to linguistics now.

And, in one episode--the same one, I think--Hoshi is trying to get the universal translator to work on a new language, and she's having trouble with it, and then says something like, "A few more words...there we go!" and the thing is suddenly able to translate the language perfectly. So the technology ends up just being more magical than ever.

[includedmiddle]

I like the idea of a bimodal grammar. It sounds like it might mean using one of two syntaxes depending on the situation. It's like having two languages in one! Or one and a half, assuming a common lexicon.

[sally]

Tensor wrote:
>>Sally, your DS9 episode "Babel" is about language! Excellent!<<

Thanks! I barely recognized it after they got to it. I had a lot more control over Hollow Pursuits. My second teleplay was about agnosia, the inability to understand speech; they bought the story only, and turned it into aphasia of the "word-salad" sort, so they could cut me out of screen credit altogether. I had to arbitrate by proving that they kept the basic premise and the title at the very least.

Tensor wrote, quoting me:
>>But: "More recent linguists talk of image-schemas dictating the very way we conceive of the universe. So EVERYTHING, basically, is a metaphor, c.f. George Lakoff, Mark Johnson, etc." Lakoff's ideas are quite controversial within linguistics. You might try Semantic Compositions' skeptical look at Don't Think of an Elephant (1 2 3 4 5).<<

Thanks for the link, Tensor. I know that Lakoff/Johnson are controversial, and that their first book (Metaphors We Live By) is simplistic; I'd be interested if you directed me to other more developed examinations of the image schema, if that is a term still in operation.

Besides putting in a bid for the episode's excellence as a story, and moving one at that, all I meant by my first post is that the problem of the Tamari language has been heavily discussed in Star Trek lists I used to frequent over a decade ago; not many of these people were linguists, but the defect is SO obvious to anyone of common sense! I, too, had considered something rather like Tamari-A and B. My surmise was that--should this idea be extended beyond the simplicities of the episode--such a language would have far FAR more (a million more!) metaphoric expressions, whose subtleties could refer more minutely to things like power converters, but that it would be like trying to understand a Malaysian commercial: you can study Malay vocabulary and grammar all you want; but you need to immersed in the culture to really speak it. (Is it Malay that I'm thinking of? I read an article about some S.E. Asian language that is so allusive in practice as to challenge foreigners... was it Hall's _Beyond Culture_?)

Now this could be the way to suggest how Tamari tricks the UT, but again, any educated person would realize that ALL languages first heard by the UT could trick it that way. How's a Klingon Universal Translator supposed to know what "beyond the pale" means in English if it has never before encountered English? Nothing would be in its data base about the difference between "pale" (stake used to cordon off Irish villages during the Irish Wars, much less anything about the Irish) and "pale" (pallid), and "pail" (bucket).

I suppose what we don't have is a linguist's published take on Tamari as it appeared in "Darmok." But my immediate sense was basically Shatner's "so what?" It's just a TV show! The principle has inspired a wonderful debate, appropriate as you say to this venue. Meanwhile, it's too bad Meyers' _Aliens and Linguists_ came out in 1980. It should be updated. A new book by Conley and Cain (Encyclopedia of Fictional and Fantastic Languages, 2006) is simply too geared towards the layman, and says nothing interesting about Tamari in its entry under the languages of Star Trek (p. 172).

Sally

[Sally]

>>Sally: the Sally Caves of Teonaht fame? Why, I know you from CONLANG!<<

Why, and I know you, too, J.S.Bangs, from CONLANG! :)

[Sally]

Does anybody have a hard time reading the code that is required before posting? Is something approximating it allowed?

My last was something that I converted into a fine new Teonaht word: ezbekomde: /Ez'bEkumdE/.

This means "Elephants floating with Mercutio upon the wind." Which means "watch out for spam."

;)
Sally

David (greetings!)

"The Samaritan Snare" is a very funny episode! Taking in your wonderfully snide remark below, I'm more convinced than ever that the clever UT was basically abandoned by Star Trek early on, and is an invisible and slightly inconsequential deus ex machina, tepidly used to contradict what we all know: the universe speaks American English in all its dialects. Except for Picard. Who speaks French. :)

Do you remember what episode on Deep Space Nine had the Ferengi traveling back in time to twentieth-century America and hearing only gibberish? Proves that Fedspeak is NOT English! (Do you think French has stayed pure into the twenty-fourth century?)

"Minuet, Data, and Jean Luc. On the holodeck."

(That means: "We will always have Paris.")

Sally

David wrote:
>>A similar question could be asked about the episode where the Enterprise encounters a crew of "dumb" aliens. Here, the UT is apparently able to decipher the language the aliens are using, and tell that they aren't using it very well. That's a pretty clever machine.<<

[Anatol Stefanowitsch]

The Tensor wrote:

It's interesting how the first few [comments] seem generally to agree with what I've written, then there's a wave of more critical comments. I'll bet that represents first the linguistically-oriented regulars, then people following links from more science-fiction-oriented sites.

Um, no. As far as I can tell, at least Sally and I have all the academic credentials necessary to call ourselves "linguistically oriented". My impression is that the first few comments agree that your analysis is correct and exhaustive, while the next few comments also agree that your analysis is correct (I certainly do) but that it fails to capture part of what the episode is "really" about.

In this context, Michael Farris posted an interesting thought over at languagehat:

Actually the Darmok _language_ (as a language) seemed pretty amenable to the UT. But Darmok ideas about appropriate modes of communication was what stymied the mostly human crew of the Enterprise.

Similarly, David Peterson comments above:

What I always took from this episode is that the Tamarians, like Lugones's chimps, had a language, but refused to use it.

This actually provides a fruitful way of thinking about this episode, and one that removes many (if not all) of the contradictions.

Tensor, in your original post you sketch out a scenario with two variants of Tamarese, Tamarese-A, which is fully compositional, and Tamarese-B, which is the non-compositional variant spoken in the episode. You then rightly point out:

But if that's the case, the whole episode falls down -- after one or two attempts with Tamarese-B, wouldn't the captain give Tamarese-A a try? Worse, when Picard says things in English to the Tamarian captain, the Universal Translator ought to be translating them into Tamarese-A. The Children of Tama, therefore, must somehow use Tamarese-B exclusively, without realizing it's encoded in Tamarese-A.

Michael's and David's postings open up another possibility: that the Tamarians realize full well that there are two variants of their language, but that they consider Tamarese-A to be completely inappropriate for 'true' communication and that they use it exclusively for menial tasks such as operating space ships. For 'true' communication, they use Tamarese-B.

Now, note that the Tamarians do not have to communicate with the Federation for any particular reason. They arrange the meeting and beam the two captains to El-Aldrel IV simply because they want to communicate. Since this is the case, surely they want to achieve what they consider to be 'true' communication, and for this they have to use Tamarese-B. In this scenario, the Tamarian captain's initial frustration would be due to the fact that he cannot get Picard to communicate in Tamarese-B.

Does anybody know of real-world historical situations where communication between cultures failed due to the fact that they had different ideas about appropriate communication rather than that they literally did not understand what the other was saying?

Sally wrote:

"Minuet, Data, and Jean Luc. On the holodeck."
(That means: "We will always have Paris.")

I love it! Proves that Tamarese is perfectly able to extend to new situations, provided speaker and hearer share the relevant cultural background.

[Sally]

I meant: "Minuet, Riker and Picard: on the Holodeck." Data and Minuet exchange a few words in French, but it's RIKER who falls in love with the Holo-Minuet.

Thanks, Anatol; yes, I'm linguistically oriented, having taken (but not majored in) linguistics courses at Berkeley, the rest self-taught, especially through CONLANG. Main interest was performatives then.

About Tamarian/Tamarese: could an interpretation of the episode be that the Children of Tama regularly seek out foreign cultures and submit them to this test of communication and bonding? The test would be to see 1) whether a foreign culture would make the effort to learn their difficult language and 2) learn it enough to understand what they want from them: a blood brothership brought about by fighting a common enemy, the monster on the planet. The point is not to kill the monster, but to work together to create a meaningful social connection. Surely other starships from the many worlds have passed through their space. These people have high standards about honor, empathy, intelligence, problem-solving and cultural interaction, and make you live up to them.

That would explain their not using Tamarian A (or was it B?). It's a test.

Sally

[Suzette Haden Elgin]

Thanks for an excellent post. Anything else I might say will already have been said in the [also excellent] comments, so I won't clutter your blog with redundancies.

[Christopher Davis]

The James White story is "Tableau", which takes place before Sector General and (via "Accident") explains the founding of the hospital. The first-contact-gone-wrong is backstory, but "Tableau" describes another contact, which eventually leads to better results.

Paul Winfield played two starship captains, neither of whom made it out of their first appearance alive. At least Dathon didn't get a worm in his ear.

[Stephen M (Ethesis)]

"Starfleet knowing their myths" ... that is the easy part and the only believable one. That a Starfleet database might have information gained from outside Starfleet is no more unlikely than the revamped ARPA net might have information from Communist sources or other places outside the U.S.

[Tweeki]

Bravo! I have always winced at this episode... Now I can point to your analysis if more depth is needed in an explanation.

The problems I have with this episode boil down to:

A major weakness of TV involving multi species first contact is that you can NOT take the time to have everybody learn each others language... Not even if you show it as peeling pages off of a calendar - because the characters are going to be really old by the time they manage to talk to a reasonable (by dramatic standards) number of species.

So, one must have a gimmick to get past this problem. As with so many other details in a SF universe (FTL, time travel, godlike power), if too much attention is paid to it --> POOF! It maksa no sense!

This episode doesn't just pay too much attention, it revolves around it. And beats it to death.

Add in fairly unattractive make up (instead picture the species looking more like Jeri Ryan "7 of 9" than Porky Pig) and the line "Chaka when the Walls fell" gets old real fast.

The way to deal with this type of gimmick is either ignore it (paraphrase of SG-1's official answer: No, the aliens all speak different languages. Why would you think they all speak English?) or treat it as background the way ST TOS did.

Oddly enough, I met a woman who felt that this was the best ST TNG episode. She liked it because it dealt with language, and she never noticed the problems.

Oh, and one more thought on the idea of languages based on metaphors: Perhaps the Tamarese have a genetic memory that includes their entire history / mythos. I.e. they don't have to learn Tamarese-b. It is built in.

[Gilgahesher]

Hi, so having not taken more than a few undergrad linguistics courses (unfortunately, at least two with Lakoff [affable prof, but the lectures weren't very good]), this episode never bothered me for its lack of plausibility. In fact, I found the fake-language surprisingly robust (given that its entirety is packed into maybe 10 mins of a 45 min show), and have busted it out on other TNG fans, and was able to "converse" (not much of a conversation, and probably says more about the frivolous and repetitive nature of online discourse than how well-packed the fake-language is, but whatever).

But hey, if the inconsistencies need explaining away:

-The UT revisionism is obviously a result of all that messy time traveling. Ha!
-Children of Tama do "think different", even more so than living rock, yep -- could happen, why not? There are all sorts of real human aphasics that process and produce languages in wacky, zany ways.
So why can't the UT be "fooled" if that's the way they actually think?
-The Darmok folklore could be common to multiple species, or ancestral -- that's how the proper names ended up in a fed database.
(The Tamarians are probably bored talking about Darmok with their non-weird alien peers all the time so this is why they seek out new pals)
-It's not necessary to bootstrap yourself with a Tamarese A to acquire the background folklore. Picard seemed able to do it just fine by context clues.
Note that when Picard finally "gets it", and starts chatting w/ the "Tamarian Riker", his utterences don't get babelfished by the UT into his literal meaning, which is consistent, because this referential Tama-mode of thought/speech isn't part of the UT's magic. This sort of constant referencing is a bit perverse and distinct from the everyday idioms and metaphors of English, for example. Imagine if you started conversing entirely in Simpsons references (the show is a large enough corpus to cover nearly every conceivable situation), and in fact the Simpsons has so pervaded culture, that you might even be able to get away with it for awhile before anyone realized what you were doing. Haw-haw! But anyways, the UT would have the same problems with Simpsonese. "Bart and Milhouse at the Quik-E-Mart!"
"Snowball! When the cat died" These are distinct from actual simpson catch-phrases that have entered the language, like "DOH!" (probably requires no translation) and "cheese-eating surrender monkeys" and whatnot.

If anything, it's this episode that should be considered canon w/r/t UT limitations and the rest of the series that should be held into question. How would a UT handle mini-Tamar-like problems that do exist in English like, "We macguyvered the warp drive together with a broken phaser and a toothbrush." I'd like to see the UT somehow find the Klingon historical figure that is the equivalent of MacGuyver to handle that translation. Subbing in the Klingon equivalent for 'jerry-rigged' instead of properly handling the macguyver ref seems inelegant somehow -- you'd totally miss the toothbrush joke. Also, you'd need a Klingon equivalent for toothbrush (I know there's some kind of fang picker, but it's not the same, man) I submit that a more 'plausible' version of the UT would fail as it does in this episode, and just pass on 'MacGuyver' as a reference for Kang and Kodos to sort out.

[Aaron L.]

Obviously, not even Star Trek tries to take it's own technobabble too seriously (although they did publish "Technical Manuals" and, like the geek I am, I purchased one!). As far as getting them out of a tight dramatic spot, the UT works wonders in every episode (as does the episode The Chase in explaining why most Trek species are all humanoid), except when they try to explain it. So in terms of Star Trek in general, I'd be very hesitant to use a TOS episode explanation of TNG technology... especially when it involved mental scanning, which is rare in TNG technology.

But in regard to the episode itself, I would add the following:
- Speaking in allusions to historical events is not at all bizarre or other-worldly. We do it every day with pop-culture references (like maybe "Morpheus, when Neo saved Trinity from the exploding helicopter"!).
- I imagine a UT working very literally, like it apparently does in TNG... without regard to context, which can explain aliens speaking in odd grammatical structures, but still being basically understandable. In this instance, context is everything, and as such, assuming the UT doesn't scan brains, which is indeed ludicrous, the UT would only translate, as you said, Tamarese A, and not B.
- Tamarian children would learn their language no different than children on Earth learn their native language. There would certainly need to be history/mythology lessons interwoven in language classes, but I fail to see how it would be different from learning pop-culture/historical/mythological references today.
- A culture which has evolved to use context-specific allusions would likely not have a concept of separating speech patterns into A and B. It would be just as odd as humans thinking that an alien species should understand our culture-specific references. If you've evolved to the point of using culture-specific references in every statement, thinking outside that box is nigh impossible, at least within the 48 minutes of a Trek episode!

Great post, though. I thought the episode was one of the most original in ST just because it approaches a species from a much higher viewpoint than Trek usually does. Outside the box...

[some guy]

I'm sure I could invent a lot of treknobabble to cover the Tamarese-A/B problem. For instance suppose you're correct, and these two languages exist, but the C-of-T are only capable of *vocalizing* the B track? (or else Track A could be carried electromagnetically or something weird like that). Or recalling the suggestion that such a language would last only one generation: perhaps the Federation, over one or two hundred years, has only *seen* one generation (they could be very long lived, like tortoises!)? And in another hundred years or so, we'll encounter an entirely different system of allusions to strange things that happened in the last hundred years?

Just sayin' is all. Cheers!

[Interest, when the episode reran]

In the good old days, the "universal translator" was some old guy who lived with the tribe for years...

[Rivernator]

Great article (just googled 'darmok' after catching a rerun), by some of your assumptions seem strangely earth/human centric for an article o. sci-fi. For example, why is it so hard to conceive of a race of beings who would might rather risk death and even war than speak baby-talk? We don't blink an eye when Klingons prefer death over dishonor...

[auto scanner]

You hit on pretty much all of the objections I thought of (Starfleet knowing their myths; using allusion to say "Hey, I think that screw over there needs tightening"; the language acquisition problem), and many others, thereby saving me the need to write this up myself with less thoroughness and less eloquence.

[auto scanner]

I'm sure I could invent a lot of treknobabble to cover the Tamarese-A/B problem. For instance suppose you're correct, and these two languages exist, but the C-of-T are only capable of *vocalizing* the B track? (or else Track A could be carried electromagnetically or something weird like that). Or recalling the suggestion that such a language would last only one generation: perhaps the Federation, over one or two hundred years, has only *seen* one generation (they could be very long lived, like tortoises!)? And in another hundred years or so, we'll encounter an entirely different system of allusions to strange things that happened in the last hundred years?

[lindsay]

Adults who share cultural awareness, a sense of humour and a quick wit will often talk in metalanguage that's impenetrable to an outsider, who knows something's being communicated, but not what. Similarly in Bantu anthropology I was reading, the wisdom of the elders is contained in speech that uses everyday words, yet is impenetrable to the immature. I thought the episode was beautiful even with its flaws, and the aftervibe I get from it chills with relax and peace. lindsay x

[Tom]

The simplest explanation I can think of is that there are religious reasons that Tamarese-B is spoken--any counterintuitive and seemingly counterproductive behavior almost has to be based on religion, yes? During the campfire scene Danoth seems to have at least a rudimentary understanding of Picard's speech, as though it were a form of communication he vaguely remembered from the distant past. The specificity issue seems insurmountable, though. In any case, for all its faults, this episode has to be respected for Patrick Stewart's performance, and I suspect it was written with that in mind, not philological niceties.

[Treknobabble120]

And yet it has inspired you to write this fine piece of analysis and you forgot to mention or you did mention and then did not explain, why this episode for all its flaws works, because it does. And maybe it works not in spite of its flaws but because of them. It is fiction, and it is not hypothesis, and it is I am going to say, useful fiction.