Notes on Ethiopic Localization

There are 87 Ethiopian and 8 Eritrean languages in the collective regions where Ethiopic text is in use. Orthographic practices are numerous; often language specific, influenced by neighboring, external, and even extinct languages. This document does not attempt to address specific orthographic practices nor does it address localization with respect to input methods, national language support or other GUI issues but is intended to provide information on and example of general text formatting issues common to computer operating and word processing systems.

The task of describing formatting practices in Ethiopia is one on par with describing the shapes of clouds in Ethiopia. Like clouds Ethiopic formatting is rather hard to pin down for study and careful description. Before one could finish describing the shape it would invariably change before you.

Fortunately, and like clouds as well, formats come from all places and in a wide range of shapes which people are willing to accept with little aversion to difference. But there is commonality among these clouds over Ethiopia that we can consider. The point here is to keep in mind that at this time there are no standard conventions for formatting text in Ethiopia but rather a plethora of defacto standards for modern practices coexisting with fairly well known rules from traditional practices.

Click here to downlaod a zip archive of these pages (long file names used -get Winzip to extract!)
Click here to view this page with Unicode addresses in place of Ethiopic images.
To send comments, corrections, and suggestions for the development of this document send email to

Formatting Ethiopic Text

Ethiopic Wordspace

The Ethiopic wordspace character, U+1361 (U+1361), was a device originally used to minimize the space between words on a line while keeping the words discernible. In this way allowing scribes to maximize the use of available space on their labor intensively produced writing material, Brana. ``Hulet Neteb'' was still in strong use during the first half of the present century. The Addis Zemen newspaper stopped using the Hulet Neteb in 1942 which is a good reference point to mark the decline of the character. Hulet Neteb is oddly used more in hand written practices today than in modern typesetting, though it remains vital to the later.

Rules of Hulet Neteb:

  1. Is used between words in lieu of a blank space.
  2. Is properly centered between words though in some publishing practices and in hand written practices will adhere to the end of the word it follows.
  3. Does not follow or precede other punctuation.
  4. Usually does not follow but may precede an Ethiopic number (this is general and not strict).
  5. Does not start a new line when a line breaks, it will be the last character on the preceding line.
  6. Is not used to delimit hours, minutes and seconds in time -some typist may do this when their font does not have a colon, or when changing fonts to use a colon would be laborious.
  7. Otherwise may use ``normal'' rubber spacing rules on either side in fully justified text.
  8. Should be recognized as having both space and punctuation character classes.

In Eritrean-Tigrigna the Hulet Neteb (or ``Kelete Netbi'') is used in place of U+1363 (U+1363) in Ethiopian practices and thus takes on a different syntactic meaning. The use as a list or numeric separator is not known to the present author.

Default Question Mark

This is really more of an IM issue; In Eritrean-Tigrigna, U+1367 (U+1367), is the preferred question mark character. U+1367 is otherwise unknown in Ethiopia where an Ethiopic-ized U+007F is a must.

Other Punctuation

In keeping with Ethiopic wordspace no additional space is inserted before or after Ethiopic punctuation. In modern practices a single word space will be added after the punctuation. It should also be noted that while an Ethiopic paragraph separator, U+1368 (U+1368), is identified in the Unicode standard for Ethiopic, it is not used in modern practices.

Ethiopicized Punctuation

A characteristic trait of Ethiopic writing is that the weight of the text is heavier than for most other scripts. The apparent ``extra weight'' should also be applied to borrowed text elements from other scripts. This included western numerals and punctuation. Doing so gives the text and a natural and continuous visual flow. Not doing so becomes visually confusing. It is the prevalent practice in professional computer fonts to add this extra weight.

Additionally, and for the same esthetic benefit, the rules of curvature may be borrowed from Ethiopic elements and grafted onto the borrowed foreign symbols. Recommended examples of this practice are demonstrated in the Monotype and SIL Premier Ethiopic fonts which enjoy widespread use in government and private houses in Ethiopia.

Ethiopic Hyphenation

Ethiopic follows similar rules to English where a word may be split over two lines at a syllable. Since all Ethiopic word elements are syllables the splitting point is considered arbitrary and no hyphenation character is used.

The lack of a hyphenation character might sound alarming, many Ethiopian words are in fact compound words so practices could easily lead to ambiguity. This indeed would be so and the practice likely never would have evolved were it not for use of the Ethiopic Wordspace (U+1361) to clearly mark word boundaries. It is only recently since the decline of Ethiopic wordspace that Ethiopic hyphenation has lead to uncertain interpretation of text. Readers knowing the context of a passage have little or no cognitive exercise in reforming a broken compound word. This area is primarily a concern to text and word processing tools.

Abbreviated Text

Abbreviation rules vary only slightly from those in American-English practices. For example Ethiopia's capital city, Addis Ababa, would be abbreviated in American-English as ``A.A.'' while in Ethiopia ``U+12A0/U+12A0'' would be the most common form. Fullstop (U+002E) is commonly used in place of forward slash (U+002F) as in ``U+12A0.U+12A0'' -take care to note that unlike American-English, when fullstop is used no fullstop is applied after the terminal character of the abbreviation.

Abbreviations are very common and fairly standard in office practices, a list of the most common can be found here.

Formatting Lists

Lists in Ethiopic text are separated by Ethiopic comma, U+1363 (U+1363), followed by ASCII space. Ethiopic semicolon and colon may also be found in use as a list separator. This is a common resort of typists when the Ethiopic comma is not available in an Ethiopic font. It is also indicative of the overlapping or interchangeable roles of the punctuation as is often their perception.

Ordered Lists

Ordered list are given in Ethiopic text using the first form of an Ethiopic syllable followed generally by a "/" or ".". In example:


After the first cycle additional cycles are given by incrementing though the syllabary


and for the 3rd cycle:


and so on.

Numbers are less often used in lists though they do play a more important role in numbering chapters and sections in books and for labeling verses in Ethiopian Bibles.

Bullet Lists

The standard shapes used as bullets (circles, squares and triangles) are accepted and used in bullet lists in Ethiopia. The Ethiopic paragraph separator, U+1368 and variant U+1368 (U+1368), should be available to the composer as a bullet item as well.

Dialogue Lists

Ethiopic preface colon is most commonly found in interviews at the end of the presenters name before the dialogue passage is given:

Time:- What was your analysis of the situation then?
- - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - - - - - - -

Preface colon may also be used to terminate an item in an ordered list as per:


Otherwise the preface colon is found to take on many of the roles filled by the western colon.

Formatting Numerals

Both Ethiopic and Western numerals are in use today. Though the Ethiopic has long since been retired to a reserved use primarily for calendar dates and demarcation of sections in literature. While Western numerals are used everywhere else following western practices.

As noted previously Ethiopic numerals may serve as their own word boundaries when Hulet Neteb is in use. Like Roman numerals their Ethiopic counterparts have bars above and below that are commonly rendered as a continuous line in a numeric sequence.

Also noteworthy is that an Ethiopic ordinal system analogous to 1st, 2nd, etc. is common where for either Ethiopic or western numerals the superscript U+129B (U+129B) is used in Amharic and U+12ED (U+12ED) in Tigrigna. U+129BU+12CD (U+129BU+12CD) becomes the superscript in Amharic when the sense is definite (as in ``the first''). The same superscripts are used with fractions.


Ethiopic numerals sequences do not use commas or decimal points. Commas are used to delimit groups of three digits in western numbers and full stop is used as a decimal separator. Also true with currency, the roles of comma and full stop are often found to reverse.


The Ethiopic numerals are a set of twenty characters. The first 9 are the digits 1-9 and next 10 are the numbers 10-100 the last is the number 10,000 though in recent decades it has fallen into misuse as 1,000. Online algorithms offer an explanation for how the numbers increment.

Negative Ethiopic numerals are not used. The format for negative western numerals is simply -123.

Formatting Currency

There is no Ethiopic currency symbol for Ethiopia's monetary unit the Birr (U+1265U+122D (U+1265U+122D)). Rather, the dollar symbol is borrowed and prefixed without a space before the value. The preferred dollar glyph uses the two unbroken vertical lines crossing the uppercase ``S''. Negative currency notation appends the minus sign before the dollar without additional space -$123

No more than two places are given after the decimal. Values of less than one Birr may still be formatted as a whole value where the leading zero may or may not be present before the decimal. As in the west the alternate formatting without the decimal is not uncommon where a superscripted U+1233 (U+1233) is used in lieu of a cents sign (i.e. ¢).

Formatting Dates and Times

There are a mirade of possibilities for the formatting of dates and time in Ethiopia. The basis for date formatting comes from the Ethiopian Calendar which is a variant of the better known Julian calendar. In brief the epoch of the Ethiopian calendar is roughly 7 years and 8 months after that of the Gregorian. The calendar has 13 months, there are 12 months of 30 days and a 13th month of 5 days. There is a leap year every four years in which the 13th month will gain a 6th day. The current year is 1991 which is a leap year. ANSI C computer code for Ethiopian and Gregorian calendar conversions, with Ethiopic formatting utilities, may be found here here.

  1. Abbreviated day of week and month names are rarely practiced. Under extreme space limitations the days of the week and month names will be simply truncated to fit the available space while the truncated name remains uniquely identifiable. This means day of week names may be given by a minimum of 1 letter (the first) and months given by the first two letters in Amharic. Tigrigna truncations require two letters for the day of week name.
  2. Ethiopic numerals are not used in digital clocks for hours, minutes, seconds. English numerals are used.
  3. Ethiopic numerals are used for dates of the month and years. Realized DATE YEAR formations with respect to numerals types as per: ETHIOPIC-DIGIT-DATE  ETHIOPIC-DIGITS-YEAR
  4. Ethiopian clocks are 6 hours back. Meaning ``12 Noon'' would be ``6 AM'' and ``6 AM'' is the zero hour. Users would likely want the option to toggle between both.
  5. Each day of the month has a proper name under Orthodox Christian practices.
  6. There is no direct analog to AM and PM in Ethiopian practices. Generally the reference is given is given in the expression of a phrase. To give a binary division around the meridian (remember this is at the 6th hour of the day) the best terms would be U1320U+12CBU+1275 for AM and U+12A8U+1230U+12D3U+1275 for PM.
  7. There are direct analogs to BC (U+12D3/U+12D3) and AD (U+12D/U+121D) in the Ethiopic calendar system. They are used much more often in date formats than in Western practices.
  8. Comma does not follow the date in date formatting. Rather the word for ``day'' (U+1240U+1290 in Amharic or U+1218U+12D3U+120DU+1272 in Tigrigna) is used as shown in the template: DAY, MONTH DATE U+1240U+1290 HOUR:MIN:SEC YEAR AD

Applying the above we can demonstrate a few of the possible accepted examples of a formatted date:

U+12A5U+1211U+12F5U+1363 U+1325U+1245U+121DU+1275 5 U+1240U+1295 01:13:46 U+1372U+1371U+137AU+1369 U+12D3.U+121D A long format date using a 24-hour clock.
U+12A5U+1211U+12F5U+1363 U+1325U+1245U+121DU+1275 U+136D U+1240U+1295 01:13:46 U+1372U+1371U+137AU+1369 U+12D3.U+121D The same but the date of the week is an Ethiopic digit.
U+12A5U+1211U+12F5U+1363 U+1325U+1245U+121DU+1275 5 U+1240U+1295 01:13:46 1991 U+12D3/U+121D Our date now uses western numerals and . is replaced by / as user preference has changed.
U+12A5U+1211U+12F5U+1363 U+1325U+1245U+121DU+1275 5 U+1240U+1295 1:13:46 U+1320U+12CBU+1275 U+1372U+1371U+137AU+1369 U+12D3/U+121D Using a 12 hour clock the hour becomes ambiguous so the meridian field is used.
U+12A5U+1211U+12F5U+1363 U+1325U+1245U+121DU+1275 5 U+1240U+1295 07:13:46 AM 1991 U+12D3/U+121D The same hour under the Western reference system. This will likely be desirable to Ethiopians working outside of Ethiopia and with foreign agencies within Ethiopia. Though the clock is a 24 hour clock, ``AM'' is added in English to clarify the Western reference.
U+12A5U+1211U+12F5 U+1325U+1245U+121DU+1275 5 01:13:46 1991 The minimalist format under horizontal space limitations. Any of the fields now absent were always independently optional.



Ugghhh... This is tedious. At least 4 basic matricies are in use, labiovelars may be added to the matricies in at least 3 different ways, then things get language specific.

Stick with the Unicode layout for now.

Character Classes

In text processing it is essential to be aware of the character class one is operating on. The table below shows the most basic divisions within the Ethiopic syllabary as defined in the Unicode standard.

[U+1369-137C]Ethiopic Digits
U+1361 Space

In linguistic processing it is a necessity to be able to detect and reset the syllabic form of an Ethiopic syllable. Fortunately the modulo class (with a modulo division of 8) of the syllable's Unicode address readily reveals syllabic form of the character. It is also essential in both text and linguistic processing to detect the number of siblings a syllable may have. The following table shows this with respect to Unicode address space.

Syllable Families:
Having 7 Forms [U+1200-U+1206]
Having 8 Forms Everything in [U+1200-U+1357]
not having 7 or 12 forms :-)
Having 12 Forms [U+1240-U+124D]
Having 1 Form? The characters in [U+1358-135A] are very rarely occurring and do not find 20th century use. It is uncertain if they are best treated as a ligature or a 13th syllabic form of their base class.

Other Classes

Classes have also be constructed from the linguistic values of the written syllables. In many Ethiopian languages a syllabic set will duplicate the phonemes of another set, or two forms within a set may share a phoneme. It is useful in linguistic processing and IM to be aware of this. Such occurrences are language sensitive and will be addressed in this document in a future revision.

These papers on regular expressions for Ethiopic also addresses Ethiopic character classes.