The upper-case phoneme

« previous post | next post »

I'm a fan of Ian M. Banks' Culture novels, but I'd like to suggest, respectfully, that they might be improved in their approach to matters linguistic. As an example, on p. 470 of his recently-released novel Matter, we learn that "Marain, the Culture's language, had a phoneme to denote upper case".

Linguists would usually call a unit that denotes something a morpheme (or perhaps a word), not a phoneme, even if it was only one phoneme long. (In fact, we sometimes find meaningful units whose effect on pronunciation is just a single feature.)

In addition, it's odd to find a morpheme that signals something essentially in the realm of writing, like alphabetic case; and also to find that Marain still uses upper case in (some of) the same ways that English does.

Some context may help to clarify why this passage puzzled me. Djan Seriy, an agent of Special Circumstances, is trying to get back to her home, the 8th level of the shellworld Sursamen, to deal with the death of her father. Traveling on the Morthanveld Great Ship Inspiral, Coalescence, Ringdown, she visits the 303rd Aliens' Lounge, where she meets an attractive stranger.

"And you," she said, remembering to be polite. "Where are you from? May I ask your Full Name?"

"Astle-Chulinisa Klatsli LP Quike dam Uast."

"LP?" she said. "The letter L and P?"

"The letters L and P," he confirmed, with a small nod and a mischievous smile.

"Do they stand for something?"

"They do. But it's a secret."

She looked at him doubtfully.

He laughed, spread his arms. "I'm well traveled, Ms Seriy; a Wanderer. I am older than I look, I have met many people and given and shared and received many things. […] I am not the person I was when I took my Full Name, and I am not definable just by that any more. A nested mystery in the centre of my name is no more than I deserve. Trust me."

Dyan Seriy thought about this. He had called himself a Wanderer (they were talking Marain, the Culture's language; it had a phoneme to denote upper case).

So, does this "phoneme" occur at the start of every sentence, and in the initial letters of proper names, as well as in certain words used as terms of art, like "Wanderer"? This passage suggests, not for the first time, that Banks is confused about the difference between a language and a writing system.

Of course, it might be that the Culture, along with its other advanced technologies, has finally achieved the goal of Enlightenment thinkers like John Wilkins, who tried to design "philosophical" languages with a one-to-one correspondence between sounds and letters, and in which the sequence of sounds or letters in each word would encode a sort of Dewey Decimal system of ontological classification. Over the past few centuries, we've learned why this is a bad idea from a semantic point of view (see Jorge Luis Borges' 1929 essay El Idioma Analítico de John Wilkins for a form of the counter-argument) and also from a phonetic point of view (see here for some discussion).

Maybe the Culture's biological manipulations of the human stock, including more-or-less eternal life and the ability to change size, shape and sex at will, have changed the characteristics that make "duality of patterning" a good linguistic idea. However, I've never seen any indication in Banks' works that he has this sort of thing in mind. Instead, I'm afraid that he just has some form of the usual naive ideas about the structure of language and the relations between speech and writing.

Klatsli Quike, to call the handsome stranger by his non-full name, turns out to be an avatoid of the Liveware Problem, a modified Delta-class General Contact Unit. A bit later in the story, he communicates covertly with Djan Seriy by flashing patterns from a laser built into his retina,

expressed in nonary Marain, the nine-part binary base of the Culture's language.

From explanations elsewhere in Banks' works, we know that "nonary" doesn't mean that the system's coding is actually base nine, but rather that the symbols in the Marain alphabet (and presumably its phoneme inventory) can be represented as 3×3 bit patterns. The resulting set of 512 symbols is larger than alphabets (or phoneme inventories) typically are; it's a plausible size for a syllabary — but it's not at all clear, as far as I know, what the mapping to pronunciation is actually like, or what it means in that context to have "a phoneme to denote upper case". Is one of the 512 symbols devoted to this? or one of the nine bits?

Banks doesn't write "hard" science fiction, and the lack of (pseudo-)explanations for displacement and mind-reading and the like in his works doesn't bother me. The problem that I have with his linguistic inventions is not that they're mysterious, but rather that they're too specific in puzzlingly implausible ways.



42 Comments

  1. Rachel Cotterill said,

    February 1, 2009 @ 10:13 am

    I haven't read the books, so it's hard to guess what Banks was intending, but the first thing that springs to my mind is the kind of prosody which may be used in English – to indicate in speech that which would be capitalized in writing. Not so much at the beginning of sentences, or even for names, but when saying something like Winnie the Pooh: “Don't underestimate the value of Doing Nothing…”

  2. Aviatrix said,

    February 1, 2009 @ 10:17 am

    I would take "phoneme to denote upper case" as "phoneme to denote the distinction given to a printed English word by the use of upper case."

    It's a fancy way of saying "You could hear the capital letter as he said it." I think the author is slightly mocking the idea that someone could say "I am a Wanderer" and have the listener know he didn't say "I am a wanderer."

    [(myl) Well, yes; but another author — say Ursula K. LeGuin — might have written "… had an affix to indicate when a word is used as a term of art", or something like that. ]

  3. Dougal Stanton said,

    February 1, 2009 @ 10:38 am

    To back up what Aviatrix says, Banks has mentioned in another story (possibly The State of the Art) that much of what is written is considered "translated" to English, which is presumably a great get-out clause when someone points out that his advanced alien civilisation "wouldn't speak like that". :-)

    [(myl) It's obvious that Marain isn't English, and neither are the many other languages that play a role in the Culture novels; and so the dialogue must be translated, just as it is in innumerable other works in many genres where the scene is set in places and times where the languages are different from the language in which the story is presented to its readers. That all should go without saying.

    What I'm commenting on is the author's meta-linguistic commentary, telling us things about the language his characters are speaking. ]

  4. bulbul said,

    February 1, 2009 @ 11:11 am

    I remember that passage, but for some reason, I thought it said "morpheme", not "phoneme". Funny.

    Dougal,
    do you mean this:

    She frequently used Marain expressions it would be impossible to render accurately into English without at least a three-dimensional diagram, and consist­ently refused to redraft or revise the text to facilitate its translation.

    I suspect that this is another example of how Banks confuses written and spoken language.

    And now for some Sapir-Whorf from The Player of Games:

    Marain was a synthetic language, designed to be phonetically and philosophically as expressive as the pan-human speech apparatus and the pan-human brain would allow. Flere-Imsaho suspected it was over-rated, but smarter minds than it had dreamt Marain up, and ten millennia later even the most rarefied and superior Minds still thought highly of the language, so it supposed it had to defer to their superior understanding. …
    Eächic was an ordinary, evolved language, with rooted assumptions which substituted sentimentality for compassion and aggression for cooperation. A comparatively innocent and sensitive soul like Gurgeh was bound to pick up some of its underlying ethical framework if he spoke it all the time.

  5. Diziet Sma said,

    February 1, 2009 @ 11:21 am

    To add to the geekery, I should point out Marian is not just around thousands of years in the future – Consider Phlebas is set in the 1400s (I think), whereas State of the Art occurs in the 1970s.

  6. MattF said,

    February 1, 2009 @ 11:46 am

    What Diziet Sma said, and State of the Art— the short story and the collection– is definitely worth reading.

  7. Boris Blagojević said,

    February 1, 2009 @ 11:58 am

    It sounds like the kind of wrong idea people often get about language, and technical terms generally.
    You do have to appreciate the effort to write at least something about the strange languages all these aliens, future humans or elves speak, but yes, sometimes it seems it would be better if the writer just kept quiet about it as usual, when he obviously hadn't checked how do things actually work.
    However, writing large-scale fiction is a huge task, and it would be impossible for a single man to do it all "right" in a lifetime. If, for example, Tolkien had been an economist instead of a philologist, he probably wouldn't have spent decades constructing languages for his world, and would fail to inform the reader about all the intricacies of the letter 'l' in his transliteration of the elvish languages (including his opinion on the probable elvish rendering of the English velarised l), but we would perhaps get a lot more information about the monetary system of Gondor, or whatever.

    My point is, you can't do it all… but you could try better: the difference between phonemes and morphemes can't be that hard.

  8. Mark A. Mandel said,

    February 1, 2009 @ 12:05 pm

    For the "upper case" part, I can easily read it as an elision for

    "it had a phonmorpheme to denote what should be written as upper case"

    Yes, this is saving the [I forget what noun goes in this idiom], or patching the crack. But if Seriy is not a linguist, this is a not unreasonable way for her to think of it.

  9. Mark A. Mandel said,

    February 1, 2009 @ 12:07 pm

    Oh, this is irritating! I used HTML in the above comment to underline "what should be written as" and to strike through "phon", and it came out all right in the preview, but it disappears in the actual display. BUG!

  10. Rob said,

    February 1, 2009 @ 12:09 pm

    My natural interpretation of the passage was the same as Aviatrix's: in Marain you can pronounce a noun in its base form or you can pronounce it as though it were a title or proper noun. I don't know enough to say whether such an accent/affix fits the definition of phoneme currently in use by linguists, but to non-specialists "phoneme" basically just means "something you pronounce".

    Does spoken language ever pick up punctuation as a component of grammar? I'm specifically thinking about the "air quotes" that seem to have become a gestural feature of spoken English—bracketing a phrase with "quote…end quote" isn't a particularly elegant alternative.

  11. Rob said,

    February 1, 2009 @ 12:36 pm

    http://www.pcplayer.de/~ps/iainbanks/artikel/marain.html provides some interesting background from Banks on Marain.

    In particular, there are only 32 primary characters in the Marain alphabet, chosen among 9-bit binary numbers so that the 3×3 representation is unique even when rotate or mirrored. According to Banks, "The rotated versions of these are generally used to represent phonemes close to the original, unrotated sound, though others have little in common with the sound of the original, being used to stand for different vocalisations." Presumably one of the rotations of a character is a type of capitalization.

    This doesn't add much to Banks' casual conflation of "phoneme", "morpheme", and "character", however.

  12. Peter Metcalfe said,

    February 1, 2009 @ 1:22 pm

    Part of the problem is surely that Banks has lost his mojo (a messy divorce I understand) and his latest offering is somewhat devoid of point. He's recycling material (_Matter_ features yet another thick headed aristocratic society). The Algebraist was a better novel but even then it suffered from recycling (for example Machine Intelligences are persecuted, which also happened to be the cause of the Idirian War back in _Consider Phlebas_!)

  13. Nicholas Waller said,

    February 1, 2009 @ 2:17 pm

    Nitpick – it's Iain M. Banks, not Ian.

  14. John Cowan said,

    February 1, 2009 @ 2:29 pm

    Mark Mandel:

    FEATURE! Or if it's a bug, it's a bug in the browser. The preview feature uses your local browser resources to view what you typed in, so all HTML features work. A lot of those features, though, are dangerous or obnoxious (an unmatched <s> can strike out everyone's comment after yours, for example), and so they are filtered by the blog software, which is probably using a very short whitelist of acceptable elements.

  15. bulbul said,

    February 1, 2009 @ 2:45 pm

    Peter,
    I strongly disagree. The aristocratic societies and Machiavellian or perhaps Borgian power games in Matter are fundamentally different from the Azad Empire in The Player of Games. And if I remember correctly from the Appendix to Consider Phlebas, the Idiran (not Idirian) War was a religious one – the word jihad comes to mind and was even used by the erudite author of the appendix in reference to the Idirans. If you recall, the Idirans were immortal beings and believed that an immortal sould could only reside in an immortal body. The frail pan-humans of the Culture (and other mortal spieces) were in need of their leadership and instruction and thus the Idiran Empire waged war on everybody.
    Now granted, Matter is far from perfect – as many have pointed out, the pacing is uneven and we could have done without some hundred pages of the Medieval Fantasy plot and I missed some of Banks' spectacular action sequences like those in Consider Phlebas that rival anything Michael Bay has ever come up with. But considering the Shellworlds and their story, the Morthanveld and the Optimae politics and the final hundred pages, to say that Banks has lost his mojo is simply unfair.

  16. bulbul said,

    February 1, 2009 @ 3:03 pm

    So, does this "phoneme" occur at the start of every sentence, and in the initial letters of proper names, as well as in certain words used as terms of art, like "Wanderer"?
    Now that I think about it, I believe what Banks has in mind is only the capitalization of nouns indicating a qualitative difference, as in "literature with capital L", "change with capital C", "ego with capital E" and so forth. Thus a 'wanderer' is a word for someone who travels a lot, while a 'Wanderer' is a designation of a certain type of person, a certain frame of mind or a certain set of qualities.

  17. Mike Keesey said,

    February 1, 2009 @ 3:13 pm

    From Wodehouse's A Slice of Life:

        "Sir Jasper Finch-Farrowmere?" said Wilfred.

        "ffinch-ffarrowmere," corrected the visitor, his sensitive ears detecting the capital letters.

        "Ah, yes. You spell it with two small f's?"

        "Four small f's."

  18. David Marjanović said,

    February 1, 2009 @ 3:18 pm

    The preview feature uses your local browser resources to view what you typed in, so all HTML features work. A lot of those features, though, are dangerous or obnoxious (an unmatched <s> can strike out everyone's comment after yours, for example), and so they are filtered by the blog software, which is probably using a very short whitelist of acceptable elements.

    Then there should be a list of approved tags somewhere where we can read it.

    Or an automatic warning "this comment still contains open tags; close them" when we click "submit".

  19. mollymooly said,

    February 1, 2009 @ 3:47 pm

    Maybe "language has a phoneme to denote upper case" is simply backwards for "writing system uses case distinction to represent (some) phonemic distinction".

  20. dveej said,

    February 1, 2009 @ 4:00 pm

    Someone should send this post and its comments to Mr. Banks, to give him a chance to comment.

  21. Dominic said,

    February 1, 2009 @ 4:36 pm

    @Rob: re: punctuation in speech

    Of course this is an isolated example, but there is the wonderful moment in Trainspotting when Ewan McGregor, in voiceover, says "Dot. Dot. Dot." as part of a monologue.

  22. Tim said,

    February 1, 2009 @ 7:34 pm

    In Gregory Maguire's Wicked and its sequels, talking animals are called "Animals". The capitalization extends to the names of types of Animal, e.g. Goat, Lion, Cat, etc. It's indicated early on that there is a difference in pronunciation between the capitalized and non-capitalized versions of the words. (He never says what the difference is, of course.)

  23. Ray Girvan said,

    February 1, 2009 @ 8:19 pm

    "LP?" she said. "The letter L and P?"
    "The letters L and P," he confirmed, with a small nod and a mischievous smile.
    "Do they stand for something?"
    "They do. But it's a secret."

    Likely Plotcoupon?

    [(myl) True enough. Specifically, Liveware Problem, the ship he's an "avatoid" of.]

  24. John S. Wilkins said,

    February 1, 2009 @ 9:49 pm

    My namesake is badly served by the Borges essay, who had never read him and went by a very bad article in the Encyclopedia Brittanica. He is much more interesting than Borges thought (and the so-called Celestial Encyclopedia is also made up).

  25. Peter Metcalfe said,

    February 1, 2009 @ 10:43 pm

    Bulbul

    The aristocratic societies and Machiavellian or perhaps Borgian power games in Matter are fundamentally different from the Azad Empire in The Player of Games.

    Except that Banks wrote more aristocratic societies than just the Empire of Azad. There is _Inversions_ and the non-culture _Feersum Enjiin_.

    And if I remember correctly from the Appendix to Consider Phlebas, the Idiran (not Idirian) War was a religious one – the word jihad comes to mind and was even used by the erudite author of the appendix in reference to the Idirans.

    Since the conclusion of the Idirian war (I declense whatever I feel like) resulted in their Minds being liberated from their shackles, I stand by my assessment. That the war was religious in tone ("Idolatory is worse than carnage" being quoted at the beginning) does not detract from my observation.

    But considering the Shellworlds and their story, the Morthanveld and the Optimae politics and the final hundred pages, to say that Banks has lost his mojo is simply unfair.

    The shellworlds as a setting is interesting but compared with what else takes place, it was the only interesting thing in the novel. The final one hundred pages was crash bang action in desperate need of some point. The Optimae politics contradicts whatever has been said about the Culture defying the norms of galactic society in previous culture novels to the extent they have become yet another caste (but then the culture novels have always had contradictions). As things stand, _Matter_ is better than _Dune_ but Banks is capable of so much better and I have the strong feeling he was writing _Matter_ solely to pay the alimony.

  26. Matt said,

    February 2, 2009 @ 6:09 am

    Don't we have a few sets of phoneme substitutions in English that signal class and social background without constituting morphemes which we call accents? I'd have guessed that the statement in the text was Banks way of suggesting that what an upper class accent does in English, Marain does with a single phoneme substitution.

  27. bulbul said,

    February 2, 2009 @ 7:35 am

    Peter,
    There is _Inversions_ and the non-culture _Feersum Enjiin_.
    True, but in my opinion, neither is really comparable. In Inversions, you have a post-Empire political landscape, in Matter, it's the opposite, one small player trying to become the top dog. I haven't read Feersum Endjinn (though I plan to).

    (I declense whatever I feel like)
    By all means, I recommend starting with verbs. You do realize that the Idiran-Idirian thing has nothing to do with declension and that the former is the form used by Banks, don't you?

    That the war was religious in tone… does not detract from my observation.
    Your observation (my emphasis): Machine Intelligences are persecuted, which also happened to be the cause of the Idirian War. I have pointed out that this is not correct. The liberation of the Idiran AI was a side effect of the Culture's victory, not the reason for fighting.

    To the admins and other participants: sorry for the OT, I shall smack myself upside the head right away.

    Matt,
    but those are just what you said, substitutions, i.e. relative shifts. Banks is referring to a single phoneme used to mark, well, something. And the problem is that we don't know what that something is, though I'm pretty sure it's not social status.

  28. Nick Lamb said,

    February 2, 2009 @ 7:45 am

    Peter, the Idiran war was a big deal because of the Homomda. The Idirans alone represented only a minor threat, and resources (initially mercenaries and other proxy forces) had been allocated to the problem accordingly. However the Homomda saw the Culture as a potential future threat and wanted to restrict its growth and influence. So they provided logistical and technology support to Idir's war machine, while retaining their diplomatic link to the Culture to negotiate an eventual peace. Thus, although the Idirans had a religious-philosophical justification for their war, the actual significance is the Homomda involvement which was purely pragmatic.

    Once the Homomda realised that the war was hopeless and that the Idirans wouldn't surrender, they walked away from it, and the rest of the war from the Culture point of view was just mop-up by machines. As to the effectorising of the Idir home world AI which "freed" it. I'd consider that this is somewhat of a euphemism, the Culture vessel which uses its effectors merely to read minds is called by others "Meat Fucker", so using your effectors to "free" an enemy AI is merely a nicer version of what's done in Excession (by a Torturer class ROU no less) to drive the Attitude Adjuster to destroy itself. Think "Clockwork Orange" not "Moses".

  29. [links] Link salad saw a Deadhead sticker on a Cadillac | jlake.com said,

    February 2, 2009 @ 8:20 am

    […] The upper-case phoneme — Language Log takes on Iain Banks. Funny stuff, and interesting reading. […]

  30. Mihai Pomarlan said,

    February 2, 2009 @ 11:40 am

    @John Wilkins:

    Care to prove that Borges knew ONLY of the article on Wilkins in Britannica, and read nothing else of or about the man? Just a few sentences in, Borges says that Wilkins' speculative work is worthy of interest.

    While decrying the lack of one of Wilkins' books from the Argentine National Library- the very one where Wilkins puts forth his language system unfortunately-, Borges also lists the works he used as a reference for his essay. Today, all can be located easily via google (though most do require purchase; or I didn't look far enough). To say that Borges only "went by a very bad article in the Encyclopedia Brittanica" is at best a mistake.

    And what's that yapping about the Celestial Encyclopedia not existing? Pretty soon you'll be saying there is no Uqbar either.

  31. Hal O'Brien said,

    February 2, 2009 @ 2:04 pm

    So… Characters who are not linguistic professionals speak in a way that linguistic professionals never would. Further, the consensus in the thread is that the way they do speak contains "errors" that non-professionals frequently commit.

    Yet this is held up as not being mimetic?

    Shocked, shocked, I say.

    If the characters are non-professionals yet somehow use professional jargon, this is a fair enough criticism. If they're professionals yet don't use professional jargon, again, fair enough.

    But to complain of non-professionals using a non-professional vocabulary strikes me as complaining that portraying water as wet is somehow unrealistic.

  32. Mihai Pomarlan said,

    February 2, 2009 @ 3:14 pm

    @Hal O'Brien:

    Reality is unrealistic.

    Also, there is some ambiguity on whether the person who uses "morpheme for upper-case" is the character while thinking or the author while explaining. The latter can be fairly complained about, as the author here sets himself up to be omniscient (we see no character-unreliable-narrator thing going on).

  33. Mihai Pomarlan said,

    February 2, 2009 @ 3:23 pm

    Oops, that would be "phoneme for upper case". There goes the layperson mixing up the terminology.

    I will now give back my subscription to Language Log. No refund is asked for.

    PS: since the link in my post above does not work, it is supposed to go here:

    http://tvtropes.org/pmwiki/pmwiki.php/Main/RealityIsUnrealistic

  34. KYL said,

    February 2, 2009 @ 3:39 pm

    @Tim, thanks for bringing up the Wicked example. That was what I immediately thought of as well. Since the characters there were speaking English, I kept on trying to imagine just how did they indicate this distinction of Animal vs. animal (Cat vs. cat, Goat vs. goat, etc.) in speech.

    Does anybody know how the musical version of Wicked dealt with this? Did they have everyone make airquotes whenever "Animal" was said, or was there some phonetic distinction that was consistently made?

  35. Hal O'Brien said,

    February 2, 2009 @ 3:46 pm

    @ Mihai Pomarlan: "Reality is unrealistic."

    Perhaps if one is a 24×7 devotee of dada. As Orwell said, "There are some things only intellectuals are crazy enough to believe."

    Those of us who are reality-based (regardless of how intellectual we may be), will just have to carry on as best we may.

  36. Mihai Pomarlan said,

    February 2, 2009 @ 4:26 pm

    @Hal O'Brien:

    Follow that tvtropes link. You need not be a devotee of dada to find that reality looks fake; just be a devotee of television (or movies).

  37. pjharvey said,

    February 3, 2009 @ 10:31 am

    So, does this "phoneme" occur at the start of every sentence, and in the initial letters of proper names, as well as in certain words used as terms of art, like "Wanderer"?

    You are making the unstated assumption that Marain and English share those uses of upper-case letters. Marain is clearly translated and presented as English in the books for comprehension, much as any foreign language could be, and thus follows many of the rules of English. Just as, for example, a Chinese-speaking character may have the order of his words rearranged for comprehension purposes when written in English so too is Marain structured with appropriate upper-case letters upon translation.

    It is quite possible that 'Wanderer' is the only word in the sentence, or entire paragraph, that contains an upper-case letter, which would certainly relieve the pressure on the 'phoneme' to be used so liberally, if not address your issue directly.

  38. Andrew said,

    February 3, 2009 @ 2:40 pm

    John Cowan: Misfeature! http://cleverdomain.org/blog/articles/On_Misfeatures.markdown

  39. Ken Brown said,

    February 6, 2009 @ 7:53 am

    I think the literary conceit is that spoken Marain isn't supposed to be a natural language at all. It is a representation of another kind of language.

    The real language is some sort of digital communication used by the AIs. The sounds used by the humans (for values of "human" including whatever non-human species are being depicted as if they were human) are an artificial formal representation of it.

    So if there is any confusion of written and spoken language going on, it is reversed. The original living language is the "written" one, it is what the computers do. The spoken language is the logical equivalent of our written languages.

  40. EMoon said,

    February 6, 2009 @ 8:36 pm

    If linguists are amused at the way SF writers handle invented languages, then consider that SF writers may be equally amused at the way linguists consider the process (and even the product) of writing fiction.

    The confusion of writer with character is common, but often (not always) mistaken. (Disclosing as another SF writer.)

  41. Aaron Davies said,

    February 9, 2009 @ 7:22 am

    @john cowan: yes, and an unclosed B can bold the whole page, an unclosed A can make it one big link, and an unclosed quoted href can make it unrenderable. it's a blogging engine's job to handle such things, and doing so via a whitelist containing only five (or whatever) tags is simply incompetence.

  42. Aaron Davies said,

    February 9, 2009 @ 7:24 am

    anybody read anathem? what do you think of orth?

RSS feed for comments on this post