Loading ...
Sorry, an error occurred while loading the content.

Re: Phonetic Transcription

Expand Messages
  • James Kane
    The trouble with that is that the species may hear very subtle distinctions that humans can t differentiate. I think I ve heard that human plosive consonants,
    Message 1 of 42 , May 2, 2013
    • 0 Attachment
      The trouble with that is that the species may hear very subtle distinctions that humans can't differentiate. I think I've heard that human plosive consonants, especially p and k, are very similar acoustically - yet humans can clearly hear the difference.

      Although I suppose the same can be said of an English speaker trying to hear the difference between an Australian language's four series of coronal consonants.


      James


      On 3/05/2013, at 11:55 AM, Patrick Dunn <pwdunn@...> wrote:

      > Well, it occurs to me that, no matter how alien their mouths, they're still
      > making *sounds*, and sounds have a limited number of acoustic features. If
      > we can draw a spectrogram of their sounds, we can find analogous
      > spectrograms among human speech -- I'd think. But then again . . .
      >
      > One way to test this would be to run some animal sounds through Praat and
      > draw some spectrograms, then associate those 'grams with similar speech
      > sound 'grams.
      >
      > Of course, when we do that we find that, for example, a cow mooing has four
      > formants, at about 450Hz, 1866Hz, 2920Hz, and 4620Hz, while a recording of
      > me saying "moo" [mu:] apparently has four formants at 330Hz, 900Hz, 2500Hz,
      > 3275Hz -- which clearly shows . . . um.
      >
      > Which clearly shows . . .
      >
      > Which . . . clearly shows . . . that I have finished grading finals and
      > need to find something to do with the rest of my day other than running
      > animal sounds through Praat.
      >
      >
      >
      >
      > On Thu, May 2, 2013 at 5:43 PM, Sam Stutter <samjjs89@...> wrote:
      >
      >> The issue is that, even if you produce a home-grown IPA replacement, what
      >> does each symbol actually mean? The grapheme /a/ relates to a particular
      >> set of features which can be identified in a sound which is recorded from
      >> human speech. If you zoom in sufficiently on a sound in audio-analysing
      >> software you can literally see the features which make a sound /a/ (or
      >> whatever).
      >>
      >> If you have a language with sounds not replicable by humans, what does
      >> each symbol actually mean if it does not relate to something measurable?
      >> You may as well be using graphemes arbitrarily since they don't mean
      >> anything.
      >>
      >> Given that you won't be associating each grapheme with a particular sound
      >> and the limits of the speech capabilities of your alien species aren't
      >> clear or analysable, there is nothing you can particularly do with a
      >> language except break it down into some analytical or logical system (such
      >> as Lexical Functional Grammar or similar nonsense).
      >>
      >> Let's say you have a word for "cat" and in your language you write it as G
      >> A V. Since we don't know how G A V is pronounced, the letters G A V don't
      >> actually mean anything. As such, there's not really much use writing
      >> anything other that the word "cat", or, at least, some sort of analytical
      >> function like "cat-noun-singular" which explains how your language handles
      >> that particular noun grammatically and syntactically.
      >>
      >> All I'm saying is that, unless you can accurately associate particular
      >> letters with specific patterns (of sound) then there's no point having a
      >> phonological transcription of your language nor, in fact, any record of a
      >> lexicon (given that your language will be a natural speech-primary tongue).
      >>
      >> In the end, you'll need accurate reference material to sounds you have
      >> recorded or can easily reproduce, which would fill vast textbooks with wave
      >> diagrams and the like.
      >>
      >> Or, alternatively, just say of a character "his voice was like a thousand
      >> drunk geckos running through treacle" and gloss the actual information the
      >> character is saying. It's a story - nobody reads a story for scientifically
      >> accurate field linguistics. Well, except for a few of us on the list, but
      >> you know what I mean.
      >>
      >> In short, fudge the phonology and ignore the details... or start studying
      >> very advanced audiology, entomology, wave physics and experimental
      >> phonology/speech synthesis.
      >>
      >> On 2 May 2013, at 1008 PM, Jörg Rhiemeier <joerg_rhiemeier@...> wrote:
      >>
      >>> Hallo conlangers!
      >>>
      >>> On Thursday 02 May 2013 18:56:48 H. S. Teoh wrote:
      >>>
      >>>> On Thu, May 02, 2013 at 05:51:51PM +0200, Jörg Rhiemeier wrote:
      >>>>> Hallo conlangers!
      >>>>> [...]
      >>>>> Many science fiction authors goof their alien languages that way.
      >>>>> When I read Alan Dean Foster's Humanx novels, I was perplexed by this:
      >>>>> he described the language of the vaguely insect-like Thranx as
      >>>>> consisting of "clicks and whistles", yet gave names of Thranx such as
      >>>>> _Ryozenzuzex_ which did not at all sound like that. (The species name
      >>>>> _Thranx_ itself is a case of this, too.)
      >>>>
      >>>> Maybe those vowels are voiceless? Which would approximate some kind of
      >>>> whistling sounds with various colors. ;-)
      >>>
      >>> Yes, the human-pronounceable Thranx names may just be such
      >>> approximations.
      >>>
      >>>>>> Though you probably still want to use some kind of adaptation of IPA
      >>>>>> for the sake of the rest of us, so that we have some kind of
      >>>>>> reference point to go on.
      >>>>>
      >>>>> Depends on to which degree such an adaptation is practical; it
      >>>>> may indeed rather be misleading than insightful. It makes no
      >>>>> sense mapping alien sounds to IPA symbols that show no natural
      >>>>> relationship to the alien sounds; I'd prefer an abstract
      >>>>> transcription of the alien sounds over an attempt to squeeze
      >>>>> them into an anthropocentric procrustean bed.
      >>>>
      >>>> [...]
      >>>>
      >>>> I suppose. Though that depends on whether Nicole already has some mental
      >>>> idea of what the language sounds like.
      >>>
      >>> She did talk about a monstrous phoneme inventory a few days
      >>> ago (some 200 consonants and also quite a few vowels).
      >>>
      >>>> Just because the process of
      >>>> speech production is different doesn't necessarily mean the aural
      >>>> qualities (as perceived by a human, anyway) wouldn't approximate human
      >>>> speech sounds, in which case it might be somewhat useful to use IPA
      >>>> adaptations to represent such sounds.
      >>>
      >>> True. From an acoustic standpoint, it is just a matter of
      >>> formants and such stuff, no matter what they are produced with.
      >>>
      >>>> But from her description, these
      >>>> aliens seem sufficiently different that the sound quality of their
      >>>> speech would be drastically different from ours, so it's probably best
      >>>> to invent a dedicated notation for them.
      >>>
      >>> As Patrick Dunn has remarked (see below), this may not work well
      >>> in a novel.
      >>>
      >>>> The underlying principles of IPA can still be used, of course.
      >>>> Parameters such as point-of-articulation, opening/closing of various
      >>>> parts of the vocal tract, the position and shape of the tongue (or
      >>>> tongues or analogous organ(s)) would still be applicable.
      >>>
      >>> Yes.
      >>>
      >>>> Speaking of which, I discovered just yesterday that birds don't have a
      >>>> larynx, but a syrinx, which is located where the trachea forks into the
      >>>> lungs, thus giving the possibility of producing multiple simultaneous
      >>>> sounds. Has anybody explored this feature in an avian conlang?
      >>>
      >>> I haven't seen any yet, but it would be an idea.
      >>>
      >>> On Thursday 02 May 2013 19:41:19 Patrick Dunn wrote:
      >>>
      >>>> IIRC, one of Nicole's design goals is inventing a plausible language
      >> for a
      >>>> novel. In that case, inventing a transcription system for her Yemoran
      >>>> language could be a very bad idea indeed. If I, an avid reader of
      >> science
      >>>> fiction, picked up a novel with names in it like &*#(&#(897, I would
      >>>> quickly put it back on the shelf and back away from it. That's
      >> assuming an
      >>>> editor would greenlight it for publication, which I would think
      >> extremely
      >>>> unlikely.
      >>>
      >>> Sure. A transcription that looks like an explosion in a
      >>> typesetting shop is not a good idea to have in a novel.
      >>> Readers want names they can remember in order to follow the
      >>> plot. (Though an alphanumeric designation such as R2-D2 can
      >>> still be catchy.)
      >>>
      >>>> In fact, I'm still not sure with that design goal that the language
      >> needs
      >>>> to be fleshed out very much. I would be quite happy to see something
      >> like
      >>>> "Jim shook hands -- or hand and talon -- with the Yemoran ambassador,
      >> whose
      >>>> name sounded something like Yaak, but with the weird scratchy whistles
      >> and
      >>>> hums that Jim could barely hear, let alone distinguish, mixed in."
      >> Cool, I
      >>>> would think: the Yemorans are weird, and Jim is just making do with the
      >>>> closest approximation he can find, which is conveniently something I can
      >>>> pronounce quite nicely in my head. Now I shall read on, and enjoy the
      >>>> plot, rather than get bogged down in linguistic theory.
      >>>
      >>> Indeed, that is a good way of handling this in the text to be
      >>> published. A more detailed transcription could be used in the
      >>> author's working notes, though; but in the novel, the Yemoran
      >>> could indeed go by the name Yaak, as the human characters hear
      >>> it in the plot.
      >>>
      >>>> Inventing a language for a novel is not automatically an aesthetic
      >> bonus,
      >>>> unless that language plays a very, very important part in the book: if,
      >> for
      >>>> example, Jim were a xenolinguist and the plot involved him trying to
      >>>> decipher a Yemoran cookbook.
      >>>
      >>> Yes. Most science fiction and fantasy novels work well without
      >>> detailed conlangs. You usually do not need more than a consistent
      >>> naming scheme. Of course, if you write a novel in which the
      >>> language itself plays an important role, you may want to expend
      >>> more attention on it, but still, you can write such a novel
      >>> without presenting a conlang - see for instance _Babel-17_ by
      >>> Samuel Delany, which is about a linguist deciphering an unknown
      >>> language and uncovering a conspiracy in which that language plays
      >>> an important role, but *no* samples of the language ever occur in
      >>> the text!
      >>>
      >>>> Otherwise, avoid linguistic howlers and otherwise just offer an airy
      >> wave
      >>>> of the hand.
      >>>
      >>> Right. Most readers won't care about the innards of the alien
      >>> language any more than about the fuel economics of the starship
      >>> engines ;)
      >>>
      >>> --
      >>> ... brought to you by the Weeping Elf
      >>> http://www.joerg-rhiemeier.de/Conlang/index.html
      >>> "Bêsel asa Éam, a Éam atha cvanthal a cvanth atha Éamal." - SiM 1:1
      >
      >
      >
      > --
      > Second Person, a chapbook of poetry by Patrick Dunn, is now available for
      > order from Finishing Line
      > Press<http://www.finishinglinepress.com/NewReleasesandForthcomingTitles.htm>
      > and
      > Amazon<http://www.amazon.com/Second-Person-Patrick-Dunn/dp/1599249065/ref=sr_1_2?ie=UTF8&qid=1324342341&sr=8-2>.
    • Logan Kearsley
      ... Perhaps they come from a planet in a binary (or formerly binary) system around a pulsar, which puts out quite a but of radio energy. That would pretty
      Message 42 of 42 , May 4, 2013
      • 0 Attachment
        On 4 May 2013 14:28, Dan Henry <rdanhenry@...> wrote:
        > The difficulty for bioradio communication is that communicative abilities
        > evolve to use existing sensory capabilities. Radio senses will only evolve
        > in an environment in which practically useful radio data exists fairly
        > consistently over long periods of time. It may well be that an environment
        > that would favor such senses would produce life too alien for us to
        > understand their psychology in any depth, much less imagine it from our
        > armchairs. Or not. It depends on how universal our own notions of
        > intelligence prove to be. One has a great deal of freedom in speculating on
        > alien biology/psychology, as we are extrapolating from a single reference
        > point of terrestrial evolution.

        Perhaps they come from a planet in a binary (or formerly binary)
        system around a pulsar, which puts out quite a but of radio energy.
        That would pretty alien. Or maybe there're just a lot of large storms,
        and being able to sense lightning flashes from far away is useful.

        I think we can get a bit more prosaic than that, though. Earthly
        lifeforms can produce pretty strong electrical charges, and can detect
        very faint electrical charges incidentally produced by the activity of
        other animal's nervous systems, but our nerves somewhat oddly don't
        actually work on electrical conduction. It's not too difficult to
        imagine alien life that *does* have electrically conductive nerves
        (electrically conductive biomolecules are not particularly hard to
        come by), in which case the starting and stopping of current in their
        nervous systems would produce faint incidental radio emissions which
        other animals could sense for hunting purposes.

        > If you wanted to actually work out such a system of natural radio language,
        > you'd do best to start by figuring out what the radio sense was originally
        > evolved for. Communication would probably use frequencies near those of the
        > sources of interest, but not the same frequencies, so that communications
        > and navigational (or whatever) radio did not interfere.

        I'm not sold on that. Human sonic language isn't distinguished from
        other natural sounds by being in an offset frequency range, but by
        having different frequency mixture characteristics.

        -l.
      Your message has been successfully submitted and would be delivered to recipients shortly.