The Hershey occidental glyphs are numbered is a sequence that runs, with gaps, from 1 to 3926 (or 3999 or 4000 to round it off). While their arrangement has considerable internal order, it is not related to the order of the ASCII character set. Subsequent additions have filled in some of the gaps, and (in the GNU Plotutils package [gp241]) added glyphs above 4000.
Dr. Hershey [Hershey 1967], [Wolcott & Hilsenrath 1976], Peter Holzmann [Holzmann 1986], and the GNU Plotutils [gp241] employ a number of not necessarily mutually consistent terms to describe the glyphs. I've adopted some of these here, but have omitted some, changed some, and have introduced new terms. I've done so to try to increase clarity, but I'm sure that I've just introduced confusion.
My basic premise is that the style of each range of Hershey glyphs can be defined by four terms: stroke count, size, tapering, and typeface. Given these divisions, each range may then be identified further by the character set is provides.
Note that the "ranges" identified here are not fonts. A "font" is a particular collection of glyphs, presumably in a consistent typeface, arranged to satisfy the needs of a particular character set (a definition such as this applies equally well to cast metal and digital fonts). They aren't typefaces either. These are just subdivisions of the Hershey Occidental glyph repertory by various features.
The basic division of the Hershey glyphs is stroke count: 1, 2, or 3. This describes "the number of lines which are used in parallel to obtain a variation in line thickness." [Hershey 1967, p. 3], This division is even more basic than size, and can form the basis for an overall understanding of the glyph sequence.
In "Calligraphy for Computers," [Hershey 1967, p. 3], Dr. Hershey uses the terms "uniplex," "duplex," and "multiplex" for these stroke ranges. Wolcott and Hilsenrath [Wolcott & Hilsenrath 1976, p. 4] use the terms "uniplex," "duplex," and "triplex." I'll follow that usage here. The stroke count ranges in Dr. Hershey's original layout are continuous: uniplex from 1 to 1000, duplex from 1001 to 3000, and triplex from 3001 to 4000. However, the glyphs added by Wolcott at the NBS to the uniplex range (238-284 (with gaps) and 910) contain some duplex glyphs. The additional glyphs above 4000 contain all manner of things.
As a second descriptive term, the glyphs may be divided into three sets based on their capital height (or in the case of some symbols the capital height of the alphabetic characters best associated with them).
In "Calligraphy for Computers," [Hershey 1967, 3] Dr. Hershey identifies three glyph heights: 9 raster units (used for FORTRAN and Cartographic glyphs), 13 raster units (used for "indexical lines of print"), and 21 raster units ("available for principal lines of print.") [Wolcott & Hilsenrath, p. 4] simplify this terminology into: "Cartographic," "Indexical," and "Normal." These are the terms I'll use here.
The "Cartographic" size, encompassing glyphs in the range 1 to 500 (all Uniplex, therefore), is the smallest. In the original Hershey distributions, contained only capital letters (in Roman and Greek), numerals, and a restricted set of symbols. Later distributions added an Italic typeface as well as lowercase and more symbols.
Although this range is termed "cartographic," it does not include a number of cartographic symbols included in other ranges.
In [Wolcott & Hilsenrath, p. 24-25], Wolcott (identified as the originator in [gp241]) adds symbols in the range 238-284 (with gaps) which are Normal, not Cartographic in size.
The "Indexical" size covers the range 1001 to 2000. All Indexical glyphs are Duplex, but the Duplex range extends beyond into the "Normal" size range.
The "Normal" size appears in all three stroke ranges: Uniplex (501-1000), Duplex (2001-3000), and Triplex ( 3001-4000), as well as the "NBS" glyphs and glyphs above 4000. Only the Triplex stroke range is exclusively Normal size.
This division is more difficult than the earlier ones. It might be best to eliminate this division altogether, as it doesn't contribute as much as the others. Its only use is to distinguish the duplex parallel-lined glyphs in the range 2501-2800 from duplex tapered glyphs.
In "Calligraphy for Computers," [Hershey 1967, p. 3] Dr. Hershey identifies three new names for alphabets "because they are not identical with existing alphabets":
The word simplex has been selected to describe those alphabets which are composed of lines of uniform thickness and have no serifs or flourishes. The simplex style of character is known otherwise as gothic*, sans serif, grotesk, light face, or block letter. The word complex may be applied to those alphabets which are composed of lines of variable thickness and do have serifs or flourishes. The complex style of character includes those which are known otherwise as standard, modern, boldface, or black letter...
*Only in America is the term gothic applied to this style of character. [original footnote by Dr. Hershey]
In related but slightly different terminology, [Wolcott & Hilsenrath 1976, p. 4] define "simplex," "complex," and "gothic" on the basis of the quantity of tapering done in the glyphs. Note here that "gothic" means a blackletter face, where Dr. Hershey employed it in its idiomatic American sense to mean sans serif. Wolcott & Hilsenrath partially identify the ranges of glyphs corresponding to these types in their index (C-1).
I've changed these terms completely, for three reasons.
First, "gothic" has two meanings, as noted above. The "Fraktur" and "Blackletter" typefaces (of which there are examples in the Hershey glyphs in the 3301-3800 range) are often termed "Gothic." Wolcott & Hilsenrath use the term in this sense. Yet in a completely unrelated usage, certain sans serif typefaces are (at least in traditional US usage) termed "gothic" (Dr. Hershey acknowledges this sense). I feel that the word had too many disparate meanings to be used for this particularly difficult division of these glyphs.
Second, "Simplex" and "Complex" as stylistic identifiers are easy to conflate with the stroke counts "Uniplex," "Duplex," and "Triplex" ("Simplex" with "Uniplex," especially).
Third, it seems to me that terms such as "Simplex" and "Complex" best describe characteristics of typefaces (which is how Wolcott and Hilsenrath use these terms), not styles of typefaces (which is how Dr. Hershey uses them). Yet Wolcott & Hilsenrath use the term "Gothic," which indicates a style of typeface, to describe characteristics of a typeface.
I'll replace these terms with three newly coined ones: Taper0, Taper1, and Taper2. These aren't very good terms - in fact they're ugly - but they have the advantage of sticking out enough that it's unlikely they'll be conflated with other terms. They're also ugly enough that with any luck they won't be used at all.
"Taper0" (Wolcott & Hilsenrath's "simplex") glyphs don't taper at all. All Uniplex glyphs are Taper0. Some Duplex glyphs are as well. It wouldn't be impossible to have a Triplex Taper0 glyph, but there are none in the Hershey repertory.
Most Duplex glyphs are "Taper1" (Wolcott & Hilsenrath's "complex") except some parallel-line glyphs in the range 2501-2800 which are Taper0. Triplex glyphs in the range 3000-3300) are "Taper1."
Triplex glyphs in the range 3301-4000 are "Taper2" (Wolcott & Hilsenrath's "Gothic").
The final division of the glyphs based on their visual forms (as opposed to further divisions based on the characters, numerals, and symbols they represent) is that of "typeface." Once again, there is considerable room for discussion and difference.
The Uniplex faces are about as simple as writing can be: just single lines. They have no variation in width between letter components. Generally, they are "sans serif" in the sense that there are no fiddly bits sticking out to further define the glyph (though it's easy enough to imagine a serif'd lineface typeface).
In "Calligraphy for Computers," [Hershey 1967, p. 3] Dr. Hershey calls these "simplex," but his definition of the term could imply more than "uniplex simplex." [Wolcott & Hilsenrath] do not further identify the cartographic range typeface, but call the normal range uniplex typeface "simplex." [Holzmann 1986] terms these faces "plain" in the Cartographic size, while at the normal size they're called "Roman" and "Simplex." The GNU plotutils [gp241] differ further (though theirs is the most extensive arrangement into modern fonts).
It doesn't seem right to call these faces "Roman" for two reasons. First, they're not really derived from the Roman letterforms. These letters originate in scratching the sand with a stick and have their modern echoes in the ballpoint pen. They share in the great utility of these two forms of writing, but have little to do with carved Roman capitals.
I'll sidestep the issue by introducing an invented term, "Lineface," to designate this type of single-stroke typeface. It's almost as ugly a term as "TaperX," but it shares with it the distinctiveness of novelty.
Two variants on "Lineface" need to be distinguished in order to maintain uniqueness of identifications: Lineface Italic (which is really just more of a slanted than an Italic face) and Lineface Script. (A certain logic would then insist that regular Lineface is Lineface Roman, but that obviates the entire purpose of this name; let's not go there.)
As used here, "Roman" designates a style of typeface, not a character set. (The character set employed by the Romans and much of the subsequent Western and Nonwestern world will be called "Latin.") In particular, some of the Greek and all of the Cyrillic glyphs are in a Roman typeface.
"Italic" has, more or less, it's normal meaning. Certain of the typefaces I identify as Italic really have more of the feel of a mere slanting of a Roman face, especially in the uppercase.
"Italic" is entirely different from the "Italian" typeface identified by [Wolcott & Hilsenrath] (and termed here "Lombardic," see below).
"Script" also has its conventional meaning.
In "Calligraphy for Computers" [Hershey 1967, p. 15], Dr. Hershey indicates that he adapted his Script glyphs from "a Headliner Typemaster of the Varityper Corporation."
The typeface identified by [Wolcott & Hilsenrath] as "German" is a form of Gothic or Blackletter typeface. The GNU plotutils [gp241] call it "GothicGerman."
In "Calligraphy for Computers" [Hershey 1967, p. 15], Dr. Hershey calls it "German Gothic" and indicates that it was adapted from a Fraktur in Jan Tschichold's Treasury of Alphabets and Lettering [Tschichold 1966]. An examination of this work suggests that of the several versions of Fraktur illustrated the Fraktur intended is the "Walbaum Fraktur" on page 179. Tschichold identifies this as a "late" Fraktur (dating to the period 1803-1828) by J. G. Justus Erich Walbaum (1768-1839). His description of it is charming.
I'll call it "Fraktur."
In "Calligraphy for Computers," [Hershey 1967, p. 15], Dr. Hershey identifies the typeface for the glyphs which now appear in the 3301 range as "English Gothic." He say that the source of this typeface was "a Le Roy lettering set for Old English." The GNU plotutils [gp241] reverse this and term it "GothicEnglish."
As indicated elsewhere "gothic" is an overloaded term that I wish to avoid. To an American printer, "gothic" might mean sans serif. To Wolcott & Hilsenrath, "gothic" represents not a typeface but the quantity of tapering I'm calling "Taper2" here.
I probably should simply call it "Old English," then. But to me "Old English" is a language, not a typeface (a language not written in "Blackletter" for that matter, and which depends heavily upon characters not present in this typeface).
Instead, I'll simply term it "Blackletter," as it is an example of a blackletter typeface.
In "Calligraphy for Computers," [Hershey 1967, p. 15], Dr. Hershey identifies the typeface for the glyphs which now appear in the 3801 range as "Italian Gothic" "because of its Lombardic origins." He identifies it as "an adaptation of a font of the American Type Founders Company," and cites page 785 of the famous 1923 Specimin Book and Catalog of the that company [ATF 1923]. The GNU Plotutils editor identifies this as "Missal Initials," and an examination of [ATF 1923] confirms that it is so.
[Wolcott & Hilsenrath 1976] term this face "Italian." While this is not incorrect, it is easily confusable (in name, not in appearance!) with "Italic." The GNU plotutils call it "GothicItalian."
It is similar to the face identified as "Lombardic Gothic Versal" in Frederic W. Goudy's classic The Alphabet and Elements of Lettering [Goudy 1952]. I'll call it Lombardic, after its origins.
A character set is the underlying communications entity separate from the form of the glyphs used to represent it.
The ordinary Western alphabet is generally termed "Latin" (e.g., the various ISO® "Latin" character set standards).
The decimal digits: 0, 1, 2, 3, 4, 5, 6, 7, 8 ,9
The Hershey glyphs include the standard Greek characters (but not archaic characters such as digamma). No accents are provided, but some variant characters such as the final sigma are.
I know nearly nothing about Cyrillic.
The Hershey glyphs encode a wide range of symbols, but these do not necessarily correspond to the symbols and punctuation used in ASCII or any other character set. Certain common symbols are omitted. It is perhaps a sign of the recent sharp decline of our society as we grasp more and more for less and less that only three decades ago Dr. Hershey did not find it necessary to encode the copyright and registered trademark symbols in his original repertory.
Dr. Hershey's report, "Calligraphy for Computers,"
was produced in the service of the US federal government
for the United States Navy.
It is therefore in the public domain.
Important disclaimers of warranty
and liability in the presentation of public domain material.
The rest of this page is copyright © 2003-2005 by David M. MacMillan & Rollande Krandall.
Permission is granted to copy, distribute and/or modify copyrighted portions of this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts and no Back-Cover Texts. A copy of the license is included in the file entitled "fdl.txt" (GNU Free Documentation License).
The computer programs here are also present as files (in the original distribution, at least) and are released, as indicated in those files, under the GNU General Public License Version 2 or (at your option) any later version. A copy of this license is included in the file entitled "gpl.txt" (GNU General Public License).
Note: Those portions of this document which are in the public domain, if any, may be copied freely. The distribution of these public domain portions is subject to all of the disclaimers of warranty and liability noted herein.
This work is distributed WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Free Documentation License for more details.
You should have received copies of the GNU Free Documentation License and the GNU General Public License along with this work; if not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
This work is distributed "as-is," without any warranty of any kind, expressed or implied; without even the implied warranty of merchantability or fitness for a particular purpose.
In no event will the author(s), editor(s), or publisher(s) of this work be liable to you or to any other party for damages, including but not limited to any general, special, incidental or consequential damages arising out of your use of or inability to use this work or the information contained in it, even if you have been advised of the possibility of such damages.
In no event will the author(s), editor(s), or publisher(s) of this work be liable to you or to any other party for any injury, death, disfigurement, or other personal damage arising out of your use of or inability to use this work or the information contained in it, even if you have been advised of the possibility of such injury, death, disfigurement, or other personal damage.
CircuitousRoot & circuitousroot.com are service marks of
David M. MacMillan
&
Rollande Krandall.
Other trademark recognition.
Presented originally by CircuitousRoot.SM