Greek 101, Fall 2007
Brian Lanter, Instructor


F First, see the topic Fonts in the sidebar, and download SPIonic to a folder you will remember how to find. To install the font in Windows:

1. From the Start menu, open Control Panel and double-click on Fonts.
2. From the drop-down menu under the File tab at the top, select “Install New Font.”
3. Use the Folder and Drive boxes to locate and highlight the folder where you downloaded the font file. The font name, SPIonic (True Type), will appear in the List of Fonts box.
4. Highlight SPIonic (True Type) and click “OK.” The font is now loaded and should automatically appear in the font list of your word processor.

        Typing non-Unicode Greek. Word processing in a non-Unicode ancient Greek font is merely a matter of learning the keyboard mapping, which is fairly uniform and intuitive among several fonts. You do not need keyboard mapping software and you do not need symbol sets beyond the ordinary keyboard. Learning the standard ancient Greek keyboard mapping is very convenient, if not necessary, for using some functions of the Perseus website (see the topic Ancient Greek online in the sidebar). Letters with direct correspondence between Greek and English are the same: a = α, b = β, g = γ, etc. Some letters without direct correspondence are assigned by similar overall shape: H = Η, Q = Θ, Y = Υ, w = ω. Two letters are assigned by similar shape elements: j = ς and c = ξ. The letter phi is assigned by sound: f = φ. Diacritical marks are assigned using various numeral, symbol and punctuation keys; some are arbitrary but some are intuitive, such as: / = acute accent, \ = grave accent, ) = smooth breathing and ( = rough breathing, etc. Here is a chart showing the keyboard mapping for SPIonic:

  SPIonic keyboard map

To type in Greek, merely set your font to SPIonic. You can switch between Latin and Greek fonts as easily as changing between any two Latin fonts. You will need to learn which numeral and function keys are mapped to which diacritical marks. The only other difficulty in using SPIonic is that it differentiates between diacritics for narrow letters like epsilon, iota and omicron, and for wide letters like alpha and omega (which one you use for the in-between letters eta, omicron and upsilon is a matter of what looks best in the situation). However, the font is wholly consistent in using lower case diacritic keys for narrow letters and upper case diacritic keys for wide letters. SGreek, another free non-Unicode font, has almost the same keyboard map, except that it does not differentiate diacritics for narrow and wide letters.

        Why can’t we type letters with pre-composed diacritics from an ordinary keyboard? The short answer is, there aren’t enough keys. The full Greek character set with pre-composed diacritics has 219 characters. Using the Shift, Ctrl and Alt keys makes plenty of simple combination keystrokes available, but it requires special software to map them to the Greek character set.

        A fuller answer is the teeniest bit complicated, but useful to know. A single keystroke of a character key on your keyboard is designated in word processing by a two-part numeric code. The first part specifies a standard range of symbols, and the second part specifies an individual symbol within that range. For example, when you type the letter “F” in a Latin font, your word processor records the code “0,70” where “0” is the range of ASCII symbols (the most primitive symbol set – it’s mostly just what you see on your keyboard) and “70” is an upper-case “F.” Whatever font you are using assigns a particular shape to code 0,70: in this document, it’s a Times New Roman “F.” But while the numeric code is conventional, the shape assigned is arbitrary. In Wingdings, 0,70 is a finger pointing right: F. SPIonic assigns the shape of the Greek capital phi (Φ) to the code 0,70. All you need to know is that typing an “F” produces a “Φ.” When you type “a!” in SPIonic, you get ἄ – an alpha with a smooth breathing and an acute accent – because SPIonic assigns the alpha shape to the “a” key and those two diacritics together, shifted to the left, to the key “!”. This shifting of diacritics, by the way, is a major difficulty with making Greek fonts look good and with making them universal. Typing diacritics on the fly, as in SPIonic or SGreek, is not quite as elegant as having them pre-composed in perfect alignment with letters of varying widths.

        There is a standard range of codes for Greek which contains, in addition to the plain letters, every vowel separately pre-composed with every possible diacritic combination, so no alignment is required. The code for “α” is 8,1. But “a” on the keyboard is 0,97, not 8,1. So if you want to use the Greek symbol range, you either need to enter the code 8,1 yourself (by selecting it from the symbol set), or you need special software to tell your word processor to record 8,1 whenever your keyboard says 0,97. Same goes for pre-composed diacritics: the code for ἄ [alpha pre-composed with smooth breathing and acute accent] is 8,116 and there is no standard keyboard equivalent, so you either enter it by hand or use special software which records some sequence of keystrokes as 8,116. Many popular fonts do have a Greek range, and there are free mapping programs available which work with MS Word for Windows. I find the mapping software harder to use than the dedicated Greek font, but it’s a matter of taste; I am also swayed by the fact that no free Greek mapping software works with WordPerfect, my preferred word processing software.

        Why can’t we type Unicode Greek from an ordinary keyboard? Unicode uses a different set of standard codes than your keyboard. Theoretically, a Unicode font functions seamlessly with HTML (hypertext markup language), the highly efficient instruction set for encoding web pages which browsing software can easily compile into the proper image, regardless of which browser or which operating system you use. In practice, Greek Unicode fonts have yet to accomplish this ideal. The principal is the same as word processing with the pre-composed Greek range: you type a letter, your HTML editing software records a code. For instance, the Greek capital phi in Unicode is “&#934.” Theoretically, if my HTML document says &#934, and your computer has any Unicode font available that has a Greek range (Windows comes with at least three: Arial Unicode MS, Lucida Sans Unicode and Palatino Linotype), then you should see a Φ on your browser screen.

        They do, after all, use HTML in Greece – but since 1982, modern Greek has been written with just one accent, the acute accent.1 Plain letters and letters with just the acute accent mark work pretty well in Unicode; other characters are less standardized. If my HTML document says “&#7940,” then you should see ἄ [alpha pre-composed with smooth breathing and acute accent], but that’s not very reliable. Unfortunately, Unicode is not as unified as the name implies: there are different standards for coding and compilation of fonts, and differences in operating systems and browsers that frequently cause Unicode Greek characters with multiple diacritics to appear incorrectly or not at all on someone else’s system. When you go to download a Greek text from Perseus, you will find peculiarities no matter which font you choose. I get better results when displaying and downloading Perseus texts in a non-Unicode font.

        So how can HTML and non-Unicode fonts function together at all? If I write a phi in SPIonic or any non-Unicode font, and translate the document into HTML, the HTML coding simply says to use the non-Unicode font to render a shape for the keyboard symbol “F.” A phi will only appear if the browser compiling the HTML code into an image on your computer has access to the specified non-Unicode font – hence my admonition to download and install SPIonic so you can view my Greek 101 website contents. When I create a PDF document, on the other hand, it carries the SPIonic font set “embedded” in the PDF code, so it doesn’t matter if the computer compiling the PDF image has SPIonic loaded. Embedded fonts, however, can slow the printing of PDF documents on some systems.


        1This explains why, when you look at the extended Greek character set in your word processor, the vowels pre-composed with just an acute accent come before vowels pre-composed with other accent marks or combinations of diacritics. The plain letters, acute accent and diaeresis take care of contemporary Greek; the remainder of the set is for historical Greek.