Patrick Giasson will give a talk at the next Atypl in Helsinky on the Cherokee syllabary. Maybe part of is research can be of help (?)

http://www.atypi.org/07_Helsinki/30_program/40_speakers/view_person_html...

Michael Everson is also a potentially rich source http://www.evertype.com/

Jim, is there no digital data that you can throw at a frequency-counting algorithm (of which there are a few now)? If not, I would get somebody to sit down with a page or two of typical text and do a hand count; linguistic frequencies tend to stabilize quite quickly. Plus deciding on sort counts is a deterministic problem anyway - you don't need to be very precise.

hhp

Ross Mills?

Because it takes me a few minutes to whip something like this up, I went ahead and ran some stats on some Cree texts I was able to find online. My sources were Moose Cree and Woods Cree the from languagegeek.com. In the table below, you just have a plain count of how many times each glyph occurs in the combined text. Hope this is helpful.


166 U+1472 CANADIAN SYLLABICS KA
145 U+1455 CANADIAN SYLLABICS TA
124 U+146d CANADIAN SYLLABICS KI
110 U+1483 CANADIAN SYLLABICS K
102 U+14aa CANADIAN SYLLABICS MA
99 U+14c2 CANADIAN SYLLABICS NI
97 U+1401 CANADIAN SYLLABICS E
85 U+1403 CANADIAN SYLLABICS I
80 U+1505 CANADIAN SYLLABICS S
78 U+14c7 CANADIAN SYLLABICS NA
74 U+140a CANADIAN SYLLABICS A
69 U+148b CANADIAN SYLLABICS CI
63 U+1424 CANADIAN SYLLABICS FINAL RING
61 U+1417 CANADIAN SYLLABICS WA
59 U+166e CANADIAN SYLLABICS FULL STOP
59 U+146b CANADIAN SYLLABICS KE
56 U+146f CANADIAN SYLLABICS KO
52 U+1405 CANADIAN SYLLABICS O
49 U+14a5 CANADIAN SYLLABICS MI
47 U+144e CANADIAN SYLLABICS TI
45 U+14d0 CANADIAN SYLLABICS N
41 U+1438 CANADIAN SYLLABICS PA
39 U+14c0 CANADIAN SYLLABICS NE
39 U+140e CANADIAN SYLLABICS WI
38 U+1511 CANADIAN SYLLABICS SHI
38 U+14ef CANADIAN SYLLABICS SI
37 U+1431 CANADIAN SYLLABICS PI
36 U+14d5 CANADIAN SYLLABICS LI
35 U+1450 CANADIAN SYLLABICS TO
33 U+14f4 CANADIAN SYLLABICS SA
31 U+152d CANADIAN SYLLABICS YA
31 U+147f CANADIAN SYLLABICS WEST-CREE KWA
30 U+1427 CANADIAN SYLLABICS FINAL MIDDLE DOT
29 U+1490 CANADIAN SYLLABICS CA
28 U+1426 CANADIAN SYLLABICS FINAL DOUBLE SHORT VERTICAL STROKES
27 U+1423 CANADIAN SYLLABICS FINAL RIGHT HALF RING
25 U+14a7 CANADIAN SYLLABICS MO
25 U+1466 CANADIAN SYLLABICS T
23 U+1481 CANADIAN SYLLABICS WEST-CREE KWAA
23 U+144c CANADIAN SYLLABICS TE
22 U+142f CANADIAN SYLLABICS PE
19 U+14a1 CANADIAN SYLLABICS C
18 U+14a3 CANADIAN SYLLABICS ME
17 U+14da CANADIAN SYLLABICS LA
16 U+1422 CANADIAN SYLLABICS FINAL TOP HALF RING
15 U+140d CANADIAN SYLLABICS WEST-CREE WE
14 U+1420 CANADIAN SYLLABICS FINAL GRAVE
14 U+140c CANADIAN SYLLABICS WE
12 U+14ed CANADIAN SYLLABICS SE
11 U+14f1 CANADIAN SYLLABICS SO
11 U+14c4 CANADIAN SYLLABICS NO
11 U+1456 CANADIAN SYLLABICS TAA
10 U+14ab CANADIAN SYLLABICS MAA
10 U+1418 CANADIAN SYLLABICS WEST-CREE WA
9 U+152e CANADIAN SYLLABICS YAA
9 U+14bb CANADIAN SYLLABICS M
9 U+147e CANADIAN SYLLABICS KWA
9 U+1473 CANADIAN SYLLABICS KAA
9 U+141a CANADIAN SYLLABICS WEST-CREE WAA
9 U+140f CANADIAN SYLLABICS WEST-CREE WI
8 U+1515 CANADIAN SYLLABICS SHA
8 U+14b6 CANADIAN SYLLABICS MWA
8 U+1406 CANADIAN SYLLABICS OO
6 U+15ac CANADIAN SYLLABICS TH-CREE THA
6 U+1541 CANADIAN SYLLABICS SAYISI YI
6 U+14d7 CANADIAN SYLLABICS LO
6 U+1434 CANADIAN SYLLABICS POO
6 U+141f CANADIAN SYLLABICS FINAL ACUTE
5 U+1489 CANADIAN SYLLABICS CE
5 U+1457 CANADIAN SYLLABICS TWE
4 U+14c8 CANADIAN SYLLABICS NAA
4 U+1474 CANADIAN SYLLABICS KWE
4 U+1433 CANADIAN SYLLABICS PO
4 U+140b CANADIAN SYLLABICS AA
3 U+1501 CANADIAN SYLLABICS WEST-CREE SWA
2 U+15a8 CANADIAN SYLLABICS TH-CREE THI
2 U+1526 CANADIAN SYLLABICS YE
2 U+14d3 CANADIAN SYLLABICS LE
2 U+14cb CANADIAN SYLLABICS NWA
2 U+143a CANADIAN SYLLABICS PWE
2 U+1428 CANADIAN SYLLABICS FINAL SHORT HORIZONTAL STROKE
1 U+1540 CANADIAN SYLLABICS WEST-CREE Y
1 U+1528 CANADIAN SYLLABICS YI
1 U+1513 CANADIAN SYLLABICS SHO
1 U+14f6 CANADIAN SYLLABICS SWE
1 U+14f5 CANADIAN SYLLABICS SAA
1 U+14f2 CANADIAN SYLLABICS SOO
1 U+14dc CANADIAN SYLLABICS LWE
1 U+14ca CANADIAN SYLLABICS WEST-CREE NWE
1 U+14b9 CANADIAN SYLLABICS WEST-CREE MWAA
1 U+14a8 CANADIAN SYLLABICS MOO
1 U+148d CANADIAN SYLLABICS CO
1 U+1470 CANADIAN SYLLABICS KOO
1 U+1461 CANADIAN SYLLABICS TWA
1 U+1447 CANADIAN SYLLABICS WEST-CREE PWAA
1 U+1439 CANADIAN SYLLABICS PAA


jim sent me a sample of the partially completed font so that i could make a sample setting... it's attached to the initial message if anyone wants to see the font in question.

Hrant, Sii, Raph

Thank all of you for your suggestions. I have considered that it may be necessary to sit down with a piece of text and count the character frequency, and Raph has done that, for which I am grateful.

I think I might do the same thing and take an average from his count and my own.

I really do admire the tenacity of it originator, Rev James Evans, and this was my way of paying a small personal tribute to him and his just plain cussedness. He certainly got precious little help from his superiors.

Thank you all again, and I hope there will be further interest in this type.

Jim

Beautiful work. Have you considered letting someone do an authorized digital version? It would be an indescribably vast improvement over what's available now, both commercial and free.

Finally, somebody is doing justice to a North American indigenous script.
All those existing line&circle designs are just plain pukey.

hhp

Raph, Hrant

Thanks for kind comments about the Cree.

Raph I have done digital version of part of the Cree Syllabic: enough to get me to the pattern cutting stage. I did a lot of the fill-ins at the point that I was cutting the glyphs out of bristol board (Goudy's method). I could certainly fill out the characters as a digital font.

Ross Mills has done a lot of work with the Aboriginal bodies and has done one or more fonts for them. Maybe if I talk nice to him I can get some kind of key-cap layout. He can be a hard man to find. There are a lot of characters in the set: if I recall rightly, there are 125. Parking them on a key position would be easy enough if I knew where to get the information.

Thank you all for your interest in the project. it's much more than I expected. And thank you Paul for posting this for me.

Jim

I think Ross has been on vacation. The characters should all be Unicode, and standard keyboard drivers exist so all you really need to do is map the characters correctly in FontLab.

You could compare your font with Plantagenet supplied with Mac OSX as a sanity check.

Cheers, Si

Thanks, Sii.

I'll try that.

Jim

Jim, I noticed that the pattern of contrast in one or the other of each of your paired glyphs often partly or totally emulates or paraphrases what a writing of the Reverend James Evans's shapes with a broad-nibbed pen might produce, but in the other flipped member of the pair, you often flip the contrast as well. Sometimes the result is that the paired glyph 'violates' ductal logic. Meanwhile the dots reside somewhat in a world of their own.

I am intrigued by all of this. And I don't want to imply that what you are doing with contrast is wrong. But what is the thinking behind this? How does this pattern of contrast manipulation play out in word- and text-length-contexts? Does this contrast flipping have a readability benefit or exact a cost?

It could be that the combination of a desire not to waste time plus a healthy disrespect for "ductal logic" can indeed produce an instance of "optimal design". There are entire worlds out there besides geometry and chirography.

hhp

Peter Enneson: yes, I noticed that too. The syllabary has two major symmetry groups. One, which I'll call "delta", has an internal mirror symmetry, and appears in four orientations. For example, "I", which strongly resembles a Greek capital delta, becomes "O", "E", and "A" when turned to 3, 6, and 9 o'clock respectively. The other, which I'll call "p", has no mirror symmetry (in other words, is chiral), and appears in a total of eight orientations: the four compass points times two for the mirror flips. For example, "KI", which strongly resembles a "p", becomes "NE", "KO", and "NA" in the four clockwise rotations, and becomes "KE" in left-right mirror symmetry (resembling a "q"), which in turn becomes "NO", "KA", and "NI" in clockwise rotation.

I see that Jim has taken great care to respect the ductal logic for the 90 degree rotations ("O" is beautifully different from a 90 degree rotation of "I"), but for the reflections, the implied angle of the pen also reflects. Note that 180 degree rotation leaves pen shape essentially invariant.

When putting the sounds together with the shapes, though, I observed that the logic of the syllabary is organized around reflections, not rotations. KI becomes KE with a reflection around the vertical axis, and becomes KA with a reflection around the horizontal. It also becomes KO with a reflection around both axes (in either order), which is also equivalent to a 180 degree rotation.

So it may well be that the contrast flipping violates the ductal logic but at the same time respects the syllabic logic. Jim, did you ever sketch the other way round? If so, did literate Canadians find one odd and the other natural?

Jim [re digitization]: I'm not surprised to learn that you used a Bezier digital outline as an intermediate step. Several straight-curve transitions have the abrupt curvature shifts characteristic of Bezier outlines, including TE (resembling "U") at 3-4 o'clock, and KI (resembling "p") at 5-6 o'clock, and at 9 o'clock.

If I were doing the digitization, I would be very tempted to release it as a free font. It seems to me that the most compelling argument against free fonts (that it takes away incentive for font designers to sell their work) doesn't really apply here - the pay-font market obviously isn't viable, or else somebody would have done a professional Cree digital font. (and btw, the font in the Unicode 3 book, which I'm using to look up the sounds, is just as bad as the other samples I linked above)

Jim [re averaging of letter frequencies]: actually, I'm not convinced that averaging is the best way to go. What you really want to do is shoot for worst-case usage. For example, for the glyphs which have West-Cree variants (mostly a different dot position), just averaging texts from West and Moose variants will leave you with about half the frequency for both variants. If you were going to set a book page in West-Cree, though, you'd want 100% of the frequency for the West-Cree variant, and wouldn't be upset at all about their counterparts sitting idle in the case. Similarly, for a very infrequent glyph such as SHO, you want to be ready for a page in which it appears in someone's name.

Si [re Unicode]: Yes, absolutely. A font like this wants to be encoded as Unicode, not in ASCII with a keyboard mapping.

Sorry if I've geeked out excessively here. It's kinda my nature.

> chiral

Hey, you sure you wanna use that? ;-)

> What you really want to do is shoot for worst-case usage.

But you also have to worry about casting so many sorts that will go unused. On the other hand, texts with divergent frequencies can leave you stuck... So here's an idea: take the square root of the frequencies as a guide.

hhp

Hrant and Raph

Thanks for your questions and very thorough alalysis of the calligraphy in the Cree.

You are right: there are arguments going on back and forth between the mirrored glyphs. I could say I did this to simply annoy you both, but the truth is I did the drawings quickly with a big juicy ad marker in my left-hand calligraphic way. The contradictions are because I was rushing as I often do and merely flipped some of the characters. In many cases I went back in and altered these on the 360 point printouts that I used to cut as paper patterns. Many of them were missed.

I agree with you that if this is to be a digital font it needs to be brought into line, and I will certainly do that.

The one saving grace about it as a metal font is that the terminations are so small as to not jump out too loudly. Since I think that the metal type will very likely never see the light of day outside my own workshop, I am going to leave it as is. At least it is an honest if hurried effort to execute the face with some style instead of the straignt geometric renderings I have been used to. On the other hand, now that ouy have pointed it out to me, it may irritate me so much that I will have to go back in and recut about 40 characters. Thanks a lot, fellas!

The idea of making this variant of Cree available as a free font is a good one. I have made an offer to a font company as a gift with the thought in m ind that sales would likely be slight, and that it would be a nice gesture to offer it as a free font. I almost cringe at the thought of making money on the backs of a culture that has been roundly dumped on in history. I feel quite embarrassed at the sight of a white man carving a Haida or Kwaikutl totem pole, and I want to avoid the notion that I have pinched something from the native culture.

If the opportunity comes about to make metal type available to Aboriginal groups on a gratis basis I would be thrilled to do that, less of course the cost of typemetal which is getting hard to find.

I have spent this morning fitting the glyphs to their set widths, working at the Thompson caster. The outside temperature is in the mid eighties and it got to be too much to handle, so I'm taking this break.

Well I got off the subject again. Once more thank you for your input. I will alter the inconsistencies is my calligraphy when making the rest of the digital font.

Jim

Ah, a left-handed calligrapher. A probable source for that "healthy disrespect".

> if this is to be a digital font it needs to be brought into line

I would actually think twice about that. These indigenous peoples -who so often become weighed down by Western conventions- might at least be given some freedom and individuality in the structure of their script.

Jim, consider donating it to/through SIL:
http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&id= _
I'd gladly put you in touch with Victor Gaultney.

hhp

Hrant

I may take you up on that reference. I need to wait a little and see what is happening with the contact I have made regarding the face.

Good news for me: The bookdealer to whom I have sold the patterns has not taken them away yet so I went down to the foundry and looked them over. I was so close to the project I had forgotten how thorough I have been. It looks like I have made all of the sensible alterations with a the knife when cutting out the printouts. I missed three, I think, and I can recut the patterns and the three mats. The partial font that is shown on this thread does not show the changes, so it really is disconcerting. I often make changes with a the knife when cutting the patterns; stress of curve, terminations etc., since it is less work to do on the computer. . . and I like the sponteneity that comes from working in a loose manner. There are probably bad things about working this way, but it suits my imaptient nature.

Thanks again for your constructive comments and enthusiasm.

Jim

Raph, you say: "[w]hen putting the sounds together with the shapes, though, I observed that the logic of the syllabary is organized around reflections, not rotations.

Compare Robert Bringhurst's comment in The Solid Form of Language: "In Canadian syllabics (the Cree and Inuktitut scripts, for instance), each basic symbol represents a consonant, and its orientation (facing up, down, left or right) signifies the vowel."

Before Jim's reply I too thought the contrast flipping might reflect another (perhaps syllabic) logic. And perhaps reflecting syllabic logic in a diversified contrast scripting deserves more consideration. I'm not sure how this could be organized or implemented. But I do wonder if freedom and individuality in the structures of the Cree or Inuktitut scripts might develop without an indigenous scribal tradition (as Bringhurst suggests).

Also, Hrant, remember that ductal logics also underly advanced non-western scripts.

Jim, I'll put you in touch with Ross via e-mail, and I look forward to coming out to the workshop to see the type when I am back in town for a longer period.

Regarding ductus, this is indeed a tricky issue for the Evans syllabics not only because of the rotation vs. reversal conventions but also because of the rotating triangles, which if actually written with a broad nib produce six different stem weights unless you rotate the hand. The visual form of writing systems tend to reflect the tools with which they were first written. In the case of the Evans syllabics, there was no primary 'writing' tool (except in Noordzij's broad sense of writing): this is a script that began life as type. There has been no chirographic evolution of the writing system, so far as I can tell. The few handwritten texts I have seen in it have shown no aesthetic consciousness, and I'm not aware of any indigenous 'calligraphy'. The script is prized for its cultural uniqueness, as part of the identity of the user communities, but its expression is wholly utilitarian.

When we white folk sit down with our calligraphic tradition to write the Evans syllabics, we pick up our broad nib pens because that is the tool we are most familiar with (at least since the early 20th century calligraphic revival). I did the same thing, when I started examining the Evans syllabics. But I'm not convinced that this is the best tool for writing these forms, and I think the evolution of the script, in both chirographic and type design terms, would benefit from much more experimentation with different tools (e.g. pointed brush) and with less fixed ductus (e.g. employing more rotation à la Noordzij's 'mannerist' classification).

[Martin, Si, the Cree and Cherokee writing systems are unrelated.]

> Martin, Si, the Cree and Cherokee writing systems are unrelated.]

Sorry, I thought Ross stuck the Cree glyphs in Plantagenet.

Si

Some larger questions / comments?

Why did Evans devise the syllabics? Was it a necessary invention because of the problems with representing indigenous speech sounds with existing alphabetic scripts? What are the sources for Evans' forms? Did he devise it as a notational system for use outside the typographical realm or only inside. Should / could he not have started with notational devices used in the pictogram tradition. Should his initiative have been more collaborative. Was his initiative pre-emptive. What is the meaning of indeginous in this context? Might studies of pictograph writing on buffalo robes and petroglyphs help us think through issues of shape?

It seems to me that without a writing tradition (where problems of construction / contrast get resolved together naturally with reference to readability and legibility all in one feed-forward / feed-backward loop over time) it's hard to know how to proceed in typographic terms. Perhaps this shows I am not a type-designer. In a way Jim's efforts--probably earlier efforts as well--raise the question of how to give shape to Evan's formal code.

> without a writing tradition ... it’s hard to
> know how to proceed in typographic terms.

?
Maybe it's harder in that there's nothing to ape (albeit in the wrong medium - with some people not even caring about that), but in terms of making something Good it would in fact seem much easier.

Yes, it's easier to simply follow orders; but no, it's not "real" - it "doesn't count".

hhp

John

When I did my key drawings for the Cree type, I held the "calligraphic" maker at a fixed angle and didn't manipulate the tool as I wrote each from. I discovered that this produces differing stroke widths. In my ignorance I like this and kept to it as I went through the many characters.

I can't see anything terribly wrong with this approach. To me it seemed to parallel flat brushwork which is not unknown to the Coastal cultures. I am aware that I didn't take a terribly studied, analytical approach to my characters. I just did what felt good to me. In a the final outcome, I prefer the face to have some kind of movement and variation of stroke, much more than the handful of rigid geometric attempts I've seen to date. I'm aware that evans' approach in devising the characaters was of strict monoline stroke and geometric construction, but I believe he did this as a method of keeping this new language simple and uncluttered, and I am almost certain from my own experiences in typecutting that his specially ground tool gave him this consistent weight and line quality.

I did make a couple of attempts at a present-day design over the past 20 years plus, but I kept coming back to this one.

Thanks for your comments. It's indeed nice to have more people than I had expected take an interest in the face. I'ver made a couple of tentative contacts to Aboriginal bodies; brobably the wrong ones, without any response.

In the end I don't expect to make a buck out of cutting this. It was simply a way of paying a small tribute to the determined Rev evans. I'd be pleased if sometime a group or individual in the Abriginal groups could get to cast this type for themselves and print some kind of text. With all the things governmants can think of to spend money on, it seems to me that a workshop could be set up to do the work.

You may recall that legendary BC printer Will Hudson was hired to teach the Inuit people to print up on Baffin Island. They produced at least one fine book. Maybe someday . . .

Jim

Jim, regarding: "[Evans] specially ground tool"

Can you go into a bit more detail about this. Sorry, I know next to nothing about Evans. Was this a writing implement or a tool he used to make his own type.

Hrant:

So how would you proceed?

Maybe like Jim? Something that makes sense. Not chirography. The comparison of which nicely fits a riddle a former co-worker once posed concerning the place we both worked at: What's the difference between FILL-IN-BLANK and a headless chicken? A headless chicken has a random chance of going in the right direction.

hhp

Hrant, so you wold hold the marker at a fixed angle, flip or rotate the results of each basic form to make templates for the other three in each set, 'correct' them on the fly when cutting them out, etc?

I wanted to know how you might proceed if you didn't have Jim's precedant to build on. I wanted to know how you would proceed if you had to work from scratch, with only Evan's monoline versions to refer too. Would you stick with the monoline approach, would you intoduce, experiment with, develop a contrast scheme? In other words, would you develop a novel shape protocol?

I think Jim's experiment is a very interesting one, and would still love to see how the glyphs we saw above and Jim's corrected glyphs work together in real text situations. I'm not at all convinced, as John intimated, that following fixed pen-orientation ductal logic across the board will produce a set of glyphs that click satisfactorily in real-text situations. That's why I asked about the thinking behind the 'norm violations"

My assumption, following Douglas Hofstadter is that 'norm violations' are a key to evolution (of typefaces too). When they catch, because they are 'fit' they become conventions and become the platform for new norm violations.

I would incorporate features that I think add to the style and/or functionality of the end result. I would not use a "marker", not even paraphrase one. I would incorporate stroke contrast, and would be unlikely to do much mathematical transformation. I was once asked to make a glyph for an invented script, and here's what I came up with:
http://themicrofoundry.com/other/shwah.gif

hhp

So you wouldn't proceed like Jim after all?

I just said: "Maybe like Jim?" His method struck me as making sense on a certain level, of being self-confident (hence progressive) not servile.

hhp

Enne Son

It is difficult to describe evans' ground cutting tool, particulary becuase I have never seen it in the flesh. There is a replica engraving tool and replica sundry other tools that are to be seen in photogrpahs on the internet. If you start searching by typing in "Norway House, Evans tools", you will eventually come across the pictures.

I can describe the cutting tool as being ground to the stroke width of the character. Unbelievably, evans simply cut his Glyphs into the end grain of hard oak wood (described by Robert Bringhurst as "burr oak"). Presumably since his matrices were carved with a single stroke to a prescribed near depth, this gave him a uniform width of line.

The photos go on to show that once he had cast lead into the resulting cavities of the wood, he then shaved the face of each piece of type level and to a consistant "height to paper".

It's my assumption that Evans achieved a uniform monoline becuase of the precision of the tool's width and becuase he was aiming for an absolute simplicity of form. This is of course only my theory.

Those who have followed in making the face available to current typesetting systems have pretty much followed the general flavout of Evans' work, and that makes sense. I was merely taking a departure for personal reasons, and I may be completely mistaken in doing this, but I do like the crispness and precision of the way my metal version prints. The obvious disparity in line width when letterpressed at 14 point is not glaring, and I believe there is a pleasant, softer more personal look to the type set in words.

Anyhow, it's done now, and that's my one shot at it.

If you can locate Robert Bringhurst (he has moved to yet another island, thsis time, Quadra) and I haven't been able to trak him down to thank him for the information he supplied me a couple of years back regarding character sets.

Incidentally, I gather that when evans devised his language he got by by cutting only nine glyphs, which when rotated 90 degrees did for everything. With the addition of over-dotted and differing tribal character requirements, the character "set" that I cut had come out to 125; to my chagrin, a lot of patterns and matrices to cut!

I'm sorry I can't direct anyone to the site with photos, but I am certain if I can stumble across it, you will find it.

Many thanks

Jim

Peter,

Ross has compiled some very good information on the background to Evan's invention: http://www.tiro.com/syllabics/James_Evans/Evans_bio.html

See also http://www.tiro.com/syllabics/James_Evans/Rossville/Rossville.html, which gives a pretty good idea of the material circumstances of the invention if one ignores the propane torch.

Arguably, Evan's principle design goal was material and labour economy. Rotation allowed him to cut a single matrix for multiple syllables and even to use the same piece of type for different syllables. I don't think Evans had any pretensions to being a type designer: he wanted to print some prayers and hymns in Ojibway and, later, Cree translation, and determined to make his own type simply because he didn't have any other. It seems to me quite likely that if he had a font of Latin type at Norway House he would have devised a Latin orthography. He doesn't strike me as the sort of man who would make unnecessary work for himself if there were a simpler way to achieve his goal.

I think in the absence of the organic development of a writing system by being written by large numbers of people over a long period of time, developing criteria for 'well formed writing' that is not purely functional, it is very difficult to progress to the kind of diversity of type design we see today, including design that is not chirographically influenced. When people sit down -- as James Evans sat down -- to create new shapes for writing and reading, the tendency seems to be almost always towards simplified, geometric shapes. I think the chirographic evolution stage is important, because it is through the making and remaking of the shapes, at different size, with different implements, that one first begins to really understand the forms and then understands how to manipulate them and, eventually, how to deliver them from chirographic models.

Hrant thinks it should be easier to produce good type design without a writing tradition behind it. I'm casting about for an example to support this view, but I simply can't find one. Type design as a craft seems to exist only within cultures with a strong writing tradition. This is not to say that type design necessarily needs to 'ape' chirographic forms, but it strongly suggests that there is an evolutionary process that cannot easily be skipped.

Jim, I think the difficulty with the strongly contrasted, broadnib ductus applied to the Cree syllabics is that we don't have a good understanding of readability in this script. I'm concerned about the considerable variance in the weight of vertical or near vertical strokes, and considering that some combinations in text may produce areas of greater lightness or darkness, depending on the sequence of syllables, and that this may result in reduced legibility, especially when light strokes are adjacent. The ruggedness of the near-monoline, traditional forms avoid this problem, which is why Ross' designs, though modulated, have a low contrast.

Consider, for example your U like form and its inverted variant. This convention of making the left vertical heavy and the right light is a typical typographic form for the Latin uppercase U. But the uppercase U is a very infrequent letter in almost all Latin-script languages: it typically occurs, like all uppercase letters, only at the beginning of words, and even then it is a relatively infrequent letter (more common in German than most other languages). So that fact that the right side of the U collapses visually, isn't a major legibility issue for the Latin script, but we shouldn't assume that because this convention works in the Latin script that it can be applied to Cree. How do Cree readers understand the visual identity of this syllable? What, to use Peter's term, is the role architecture of the verticals? If it is to bound a space, linked by a curved bow, then weakening one of the verticals may reduce legibility. I don't know what the answer to that question is. I don't think anyone else does either: as far as I know, very little study has been done of syllabic reading in general*, and I'm not aware of any studies of Cree or Inuktitut reading.

* The obvious place to conduct studies on syllabic reading in general -- e.g. to compare and contrast with our developing understanding of alphabetic reading -- would be Ethiopia and Eritrea, where there are very large numbers of literate adults reading a syllabic script on a daily basis. Part of the problem with designing type for Cree is that relatively few people read it, and there are few texts published in the script; the situation is better for Inuktitut.

> I simply can’t find one.

Maybe it simply hasn't been done right yet? Although Hangul is admirable in its lucid innovation. In any case, the past is never "complete" - relying on it for affirmation of what one should aim for in the future is death.

A script that has arrived into our age with no
handwritten tradition is lucky, because it is free.

hhp

A script that has arrived into our age with no handwritten tradition is lucky, because it is free.

And like most 'free' things it wanders arbitrarily without anchor, having no point of reference outside itself, essentially antisocial. You've spent too much time among the American individualists, Hrant. :)

Much of the richness of anything that humans make is in how it relates to other things that we make, how it becomes part of the fabric of culture, bound to the visual and aesthetic considerations that link creations in different media. Perhaps this will make more sense to you if I say that a script without a chirographic tradition is as handicapped as a script without a painting tradition or even a musical tradition. You see? I'm not saying that type design needs a chirographic model per se, but that it needs to be informed by something other than the whim of the type designer. On this, I think we agree theoretically at least. You want type design to reference an understanding of reading, rather than an understanding of writing. This sounds like a great plan, except for the fact that a) our understanding of reading as it relates to design is limited and b) what we do know doesn't suggest a prescriptive design agenda because we read comfortably across so many different scripts and styles. So I'm concerned that there isn't enough knowledge to anchor this reading-centric approach to type design, and it will simply be used to justify arbitrariness.

The Cree/Inuktitut syllabics are an interesting case of a script that is very important to its users in terms of 'identity politics', i.e. it is part of what makes its users culturally distinct from the dominant civilisation in this country. At the same time, though, it never seems to have been naturalised into the broader aesthetic cultures of the Cree and Inuit peoples. It is something apart from their art and craft, and has not yet found an aesthetic expression within those cultures.

> it needs to be informed by something other than the whim of the type designer.

Well of course. But please, not the wrong thing.
Limited? Yes, always. But better than misguided.

> The Cree/Inuktitut syllabics are an interesting case of a script
> that is very important to its users in terms of ‘identity politics’

Which is another reason to make it cousciously different than Latin
type, which is 99% an admixture of chirography and constructivism.

I might say that the main thing it shouldn't be, is Modernist.

hhp

Jim, thanks for the elaboration, and John, those links were very informative.

John, I think you are probably right about the importance of contrast reduction in shapes such as those Evans gave us if we are to start with a dimensional vector model (broadpen-based) of how to shape the strokes. To some extent the Evans shapes are in an inbetween world between western minuscules and majuscules. And this inbetweeness has made its way into Jim's glyphs in the left and right stem characteristics of the U-type forms. So I'd say a dimensional vector approach has its pitfalls here, probably because it has not played a part in the formation of the role architecture of the characters. And it is probably not good to sustain it in any rigorous way accross theJim says, "The obvious disparity in line width when letterpressed at 14 point is not glaring." I hope Jim will give us a peek at how these glyphs set in a real text context.

I found it very interesting to see that Evans character set is recognizably descendant from the James Frere, Thomas Lucas, and Isaac Pitman efforts, which are essentially stenographic.

It is hard for me to see at the moment where perceptual processing issues come in. Because the system is syllabic, and at least origianlly aimed at pre-literate populations do readers ever come to process entire groups of them as one perceptual unit the way we do with words?

John, I think you are absolutely right about this: "what we do know doesn’t suggest a prescriptive design agenda because we read comfortably across so many different scripts and styles." To your 'prescriptive,' I would want to add 'proscriptve.' I also think that what we know helps us understand why strategic construction, a well-motivated contrast sheme, and 'space craft' still matter. I also think that what we know about parafoveal pre-processing and cue-value should be kept within our sights and might move us in new directions. I doubt that you would contest this. But I say it because I want to emphasize that just because it's not prescriptive or proscriptve, that doesn't mean it's not relevant.

And why shouldn't a Cree Modern be available alongside a Rimmer Cree? The group Kashtan appropriated a rock idiom to great effect.

Peter

I'd like very much to be able to show on this forum Cree printed letterpress, but unfortunately I have no digital camera, scanners or zip with which to do that, not to mention I wouldn't know how to do it if I did have the stuff. I am however experienced with the use of the postage stamp and would be happy to mail you a proof, if you will give me your dirt address offline.

And, hey, I thought my version WAS a modern! (joke)

Thanks for your interest.

Jim

...do readers ever come to process entire groups of them as one perceptual unit the way we do with words?

According to the anecdotal evidence of one of Ross' main contacts in Nanavut, yes. This person is actually a non-Inuit who has learned the language and script. He notes that common sequences of syllabics are recognised as such: syllabic bouma shapes.* Of course, as we know, a person's description of how he reads doesn't necessarily conform to the reality of the optical and cognitive processes.

* Actually, now I think about this, multi-glyph bouma shapes are potentially more useful in reading syllabic scripts than they are for reading English, since the spelling is much more consistent and the correlation to spoken language much more regular and precise. This means that any repeating sequence of two or more glyphs will always correspond to exactly the same sounds, allowing variation for stress and emphasis of course.

Some of these thoughts are informed by my work on Ethiopic, a syllabic script that I know better than Cree/Inuktitut (Ross is the one who has done a lot of work for North American aboriginal languages). Interestingly, surviving examples of the south semitic Sabaean script from which the Ethiopic or Ge'ez script is evolved are very geometric, regular and sometimes monoline. This is no doubt partly due to the fact that almost all surviving texts are stone inscriptions; the Sabaeans presumably had a less rigid, more informal style for other kinds of writing. But if you compare the richness and relative variety of the long and glorious Ethiopic manuscript tradition to the absence of such a tradition in the Cree and Inuktitut you can see how the organic evolution of a writing system through writing is productive of new ideas and styles. Maybe we should give the Cree and Inuit another couple of thousand years with their script and see what they come up with.

A bouma is as large as three things in combination allow: reader experience; form distictiveness (a compound of silhouette divergence, crispness, etc.); and semantic context. The blank space is indeed not an unbreachable wall.

hhp

re: evolution of a writing system through writing

See Gerrit Noordzij's Letterletter page 157. (The arrows are missing--they should proceed from item 1 in the lower register to item 1 in the upper register, then back down to item 2 in the lower register, then up, then down, etc.). Also see The stroke of the pen page 48.

standard keyboard drivers exist

can you point me at these, sii?

Paul

You may be able to get some help from John Hudson on the layout. I know his partner Ross Mills has a lot of experience with the Syllabics.

Jim

I'm checking on the keyboard drivers. We don't ship one, but I should be able to track one down. I'm checking with the the keyboard gurus here.

In the interim... http://www.wats.ca/resources/unifiedcanadianaboriginalsyllabics/36

might be helpful

Si

for those of you interested, we are nearing the completion of this project. Jim and I have been working together on the font and there is now a full set of characters covering the Canadian Syllabics Unicode block. If anyone wants to beta test the font and provide feedback, please contact me at paul(a)p22,com.