Steno IPA dictionary

[Rob]

Using Wiktionary, I've made a dictionary that outputs IPA.

Here's a way to make it (sort of) work -- Linux only so far. In the latest Plover codebase, change the function send_string(self, s) to the following in plover/oslayer/xkeyboardcontrol.py:

    def send_string(self, s):
        """Unicode-ish output hack"""
        # shows "real" output on terminal
        sys.stdout.write(s)
        sys.stdout.flush()
 
        for char in s:
            # chosen arbitrarily
            keycode = 253
 
            ordchar = ord(char)
            # keysyms for Unicode characters
            if ordchar > 100:
                ordchar += 0x01000000
            self.display.change_keyboard_mapping(keycode, [{ordchar}])
            self._send_keycode(keycode)
            # so that keys in sequence do not overlap
            sleep(0.05)
        self.display.sync()

Obviously, a better way to do it would be to assign individual keycodes to all IPA symbols. Makes me wonder how alphabets that are bigger than 255-8 characters are handled, though -- perhaps using the keycodes as a circular buffer.

Unfortunately, some symbols such as ð do not appear as they should, but they are fixable with a search-and-replace. The output of xev hints at XmbLookupString being the culprit, but I didn't look into it further yet.

KeyPress event, serial 66, synthetic YES, window 0x3e00001,
    root 0x15d, subw 0x0, time 0, (1,1), root:(1,1),
    state 0x0, keycode 253 (keysym 0x10000f0, (no name)), same_screen YES,
    XLookupString gives 2 bytes: (c3 b0) "ð"
    XmbLookupString gives 1 bytes: (f0) "�"
    XFilterEvent returns: False

[Rob]

One thing I forgot: the following line has to be placed near the beginning of the file as well.

from time import sleep

[Mirabai Knight]

This is seriously rad. I remember early after we formed this group,
some of the linguists among us talked about using steno in linguistic
fieldwork, to document the sounds of previously undocumented languages
in realtime. This seems like an important step in making that
practical.

[Rob]

            # keysyms for Unicode characters
            if ordchar > 100:

This line should read:

if ordchar >= 0x100

which fixes the characters that go haywire.

[Snowy Wilderness]

This is awesome. I think IPA will also prove to be a useful tool in steno theory and pedagogy for non-linguists: to be able to associate the various ways of encoding/expressing each word/pheme together, so that it is recognized that each word/pheme has an IPA, an English, a Shorthand, and a Steno facet, as well as a host of sentence-structure category tags it fits into, will be an important milestone in the open source steno revolution...collecting and collating full metadata of these varieties for each word, phrase and fragment will allow us to design a vastly more user-friendly steno theory and especially a better steno pedagogy, where beginners' drill-work will be able to be more data-driven, readily associative in the context of everyday language use, and easier to integrate with basic communication tasks like corresponding and blogging from an earlier stage of study.

A real-world example of what I'm talking about is the fact that--at least in our NYCI theory--after one learns the basic steno alphabet that Mirabai and her graphic designer collaborator so beautifully laid out in Mirabai's historically monumental chart, one learns a number of essentially alphabet-like key combinations that enable the integration of sounds like "KSHUN" into words in certain situations where the basic steno alphabet alone would send one into paroxysms of syllabic/orthographic contortion in the attempt to fit certain syllables like that into the linear flow of sounds and syllables and outlines otherwise facilitated by the basic keyboard layout. We definitely need another alphabetic chart for these syllables, and having the IPA for "KSHUN" be part of the standard package of databased facets of the steno combination "-BGS" (or whatever it proves to be in Plover theory, that's just NYCI, although I like it!) will provide us with some SUPER exciting options for teaching steno and then for IPA-enlightened use of steno in the field!

Sent from my iPad

[Tony Wright]

Yes! This is so cool! You could do broad IPA transcription in real
time. Make a machine-readable corpus of spoken language. Phonologists,
phoneticians, dialecticians, sociophonologists, should learn about
this. Maybe some of us should do a presentation on this at next year's
LSA conference?

--Tony

[Snowy Wilderness]

Yes, I'd LOVE to prepare for an LSA presentation...and another thing I'd love to do, sooner rather than later, if it's feasible for a good quorum of us, is to do our own Plover Steno Conference at the art space I curate on the Lower East Side of Manhattan, Culturefix...we have the space and time to get a core group of us together to set the agenda for the distributed but coordinated work we'll then all be able to do on building the theory and pedagogy needed for the popularization and optimization of stenography in the Plover Era that is now upon us.

--

Yours Truly,

Johnny Sagan aka Snowy Wilderness

[Snowy Wilderness]

Apropos: http://andrewsullivan.thedailybeast.com/2013/01/the-best-ways-to-learn.html...on non-obvious research results on optimal learning strategies.

[Snowy Wilderness]

(the upshot is that it's all about "distributed practice" and practice testing...two techniques that stenography REALLY lends itself to, so good news for us: now let's design a better course of study secure in that knowledge!)

[Ted Morin]

Maybe it's time to revisit this -- Plover supports Unicode much better nowadays. Just searched around as I'm taking a linguistics class right now. Might be possible to make an orthographic IPA theory based off of English steno.