Niamey 1978 & Cape Town 2018: 3. Other angles on Wikipedias in extended/complex Latin

Last August, I began a set of three posts marking the 40th anniversary of the Niamey 1978 meting on harmonizing African language orthographies, and associating that with the Wikimedia 2018 conference in Cape Town - the first in sub-Saharan Africa. This post concludes the series.

The central element of this discussion is the extended Latin alphabet, which is used in the orthographies of many African languages to accurately convey meaningfully important sounds, and which often makes input with ordinary keyboards or keypads difficult.

Composite image from welcome page of Eve Wikipedia.

Looking again at Wikipedias in extended/complex Latin

So far, this article has raised some questions, made some admittedly superficial comparisons, and speculated as to factors related to success or not of Wikipedia editions in African languages. What might be ways of improving this analysis?

One way would be different approaches to sorting and categorizing the Wikipedias using the same numbers in the tables in the previous post. And one of those approaches could be to consider the relative degree of complexity of each orthography within a category. For example in category 3, Fulfulde uses 3, 4 or 5 extended Latin characters and Hausa 3 or 4 (depending on the country), and Luganda and Northern Sotho only one. Yoruba, when typed with the precomposed dot-under (aka "subdot") characters and combining diacritics for tones is less complicated than the same language using the classic style small vertical line under - because the latter requires a combining diacritic and that may mean "stacking" two diacritics on a character where a tone is involved.

A different way would be to look at the quality of content in the various Wikipedia editions. In some, the raw numbers of articles (which are the figures used in the tables mentioned above) are inflated by shell articles, by which I mean a stub that may have only a few words of text and perhaps an image. The list of Wikipedias does include a "depth" metric, which might be used (or perhaps adapted) to look for possible correlations between the quantity and quality of the content on the one hand, and the nature of the orthography on the other.

Yet another way would be to consider the numbers of people working on these editions. Wikipedia counts the numbers of users, active users, and administrators per edition. Could one use these figures to better understand whether the more successful Wikipedia editions in extended Latin (in terms of numbers of articles or a depth metric) are so because of the efforts of a relatively small number of users? That's not to imply any negative judgment of such cases, but it would be useful to know if if a complicated writing system (from the point of view of input) is not a hurdle for a large number of contributors (active users), or if it's really a case of a few savvy individuals carrying the load.¹

And another approach would be to expand the scope of analysis to consider other factors: How many people speak the language? Is it taught in schools? How much printed material is available? Are there different dialects or written conventions that a contributor to or a reader of a given African language edition of Wikipedia must navigate? Any of these and perhaps others might, individually or in combination, affect the potential production of web content in general, and the success of these Wikipedias in particular.

One could also put the numbers in the background and do a qualitative study focusing on the experience of the editors of African language editions of Wikipedia. What might emerge from such discussions concerning the range of tasks involved in building and maintaining an active Wikipedia?

And then there are some stray questions certainly worth checking out. For instance, does the base interface language on which all of the African Wikipedias are built (English vs French) have any bearing at all on the success of Wikipedia editions? What about the degree of localization of the interfaces (from English or French to the language of content? And does that degree of localization relate at all to the complexity of the script?

Research towards success with African language Wikipedias

Although the number of Wikipedias in African languages is relatively small (about 13% of all editions, and collectively contain less than 1% of the total number of articles in all Wikipedias combined²), there are arguably enough data and diverse user experiences to give us a better idea of both how to develop small Wikipedias in Africa, and how much of a factor the scripts used to write them might be in their relative success.

Looking beyond African languages to the experience of Wikipedia editions in other languages written in extended Latin (and non-Latin scripts) would be instructive. This would likely highlight not only methods to facilitate input of diverse writing systems, but also supportive environments (or "localization ecologies") for these languages in general. 

Success for African language editions of Wikipedia may not be found in imitating work on other editions so much as it would in identifying ways to leverage the strengths and unique resources of African language communities. Nevertheless, facilitating input is fundamental, relying at its most basic level on common technology (for keyboards, etc.) and features of the Mediawiki software.

With rapid advances in language technology, an additional focus should be how to adapt speech-to-text to African languages to facilitate creation of content from oral narratives, interviews, and exposition. This is a topic I hope to return to later.

---

1. The Yoruba (category 4 orthography) and Northern Sotho (category 3) Wikipedias, for instance, each benefited at different times from large numbers of articles created by a single user in their respective communities.
2. That's 38 of 292 if the tiny Dinka edition is included; 37/291 if not. And about 293k articles out of 48.6 million total. (All as of July 2018.)

Niamey 1978 & Cape Town 2018: 1. Some thoughts about extended Latin & content in African languages

Image features the 31 modified letters & diacritic combinations in
the African Reference Alphabet, 1978. (Nor all are currently in use.)

The world 40 years ago, when the Meeting of Experts on Transcription and Harmonization of African Languages took place in Niamey, and that of the Wikimania 2018 conference in Cape Town (which ended last month) seem very distant from each other. But from the angle of the written form of African languages at least, the concerns of the two events are not so distant.

One of these concerns is the extended Latin alphabets that were on the agenda in Niamey, and which are used in about half of the African language editions of Wikipedia. This post and the next consider these two vantage points, asking whether extended Latin is associated with less content creation, and what might be done to facilitate its use of the longer Latin alphabet.

Adapting the Latin script to African realities

In 1978, representatives of countries that had gained independence no more than a couple of decades earlier, or in some cases only a few years before, met in Niamey to advance work on writing systems for the first languages of the continent. One of the linguistic legacies of the colonial period was the Latin alphabet (even in lands where other systems had been used). But given the phonological requirements sometimes very different than what Latin letters represented in Europe, linguists added various modified letters, diacritics, and digraphs to write African languages (sometimes even a special system for a single publication¹.

So, that legacy also often took the form of multiple alphabets and  orthographies for a single language, reflecting the different origins of European linguists (frequently Christian missionaries from different denominations), locations in which they worked (perhaps places where speakers of a language had particular dialects or accents), and individual skills and choices. After independence, many African countries undertook to simplify this situation, but they still often ended up with alphabets and spelling conventions different from those in neighboring countries.

The linguists and language specialists in Niamey, as in other such conferences of that era (many of which, like the one in Bamako in 1966, were supported by UNESCO), were concerned with further simplifying these discrepancies, with accurate and consistent transcription of languages that were for the most part spoken in two or more countries (whose speaker communities were divided by borders). That included adopting certain modified letters and diacritic combinations for sounds that were meaningfully significant in African languages (some of which correspond with characters in the International Phonetic Alphabet).

Language standardization, which is actually a complex set of decisions, was a real concern where there were on the one hand diverse peoples grouped in each state and on the other hand limited resources for producing materials and training teachers. At its most basic level, though, standardization of any sort required an agreed upon set of symbols and conventions for transcription.²

A reference alphabet for shared orthographies

The African Reference Alphabet (ARA)³ produced by the Niamey meeting was an effort in that direction. It built on the longer post-independence process to facilitate use and development of written forms of African languages - a process that had its roots in the early introduction of the Latin script (before the formal establishment of colonial rule) and efforts during the colonial period such as the influential (at least in the British colonies) 1928 Africa Alphabet. The ARA was intended - and to some degree at least still serves - as sort of a palette from which orthographies for specific linguistic, multilingual national, and cross-border language needs could be addressed.⁴

And that set of concerns - alphabets, orthographies and spelling conventions - turned out to be the starting point for later efforts in the context of information and communication technology (ICT) to localize software and interfaces, including Wikipedia and other Wikimedia interfaces, and to develop African language content online, including for Wikimedia projects. Even if it does not seem as visible as other challenges.

What I haven't seen is an evaluation of the efforts at Niamey and the other expert meetings on harmonization of transcriptions, although the most used of the characters in the ARA can be seen in various publications, and all but perhaps one are in the Unicode standard.

In any event. the situations of the various African languages are diverse, with some having well established corpora while others are "less-resourced," and in the worst case, inconsistently written.

Extended Latin and composing on digital devices

One important element in discussions in the process of which Niamey was part, was the role of modified letters - what are now called extended Latin characters - in transcribing many African languages. The ARA includes no less than 30 of them (22 modified letters and 8 basic Latin with diacritics⁵). These added characters and combinations are not intended to all be used in any one language, but represent standard options for orthographies. The incorporation of some of these into a writing of a single language makes the writing clearer, and has no drawbacks for teaching, learning, reading, or handwriting (although there are arguments against the use of diacritics). Since the establishment of Unicode for character encoding, the screen display of these characters is not a problem (so long as fonts have been created including glyphs for the characters).

However even the presence of even just one or two extended Latin characters leads to problems with standard keyboards and keypads - where are you going to place an additional character, and how is the user to know how to find it? This is a set of issues that was of course recognized back in the era of typewriters. One of the spinoffs from the Niamey conference was the 1982 proposal by Michael Mann and David Dalby (who attended the meeting) for an all lower-case "international niamey keyboard," which put all the modified characters (of an expanded version of the ARA) in the spots normally occupied by upper-case letters.

While that proposal never went far (I hope to return to the subject later) - due in large part to its abandonment of capital letters - it was but one extreme approach to a conundrum that is still with us. That is, how to facilitate input of Latin characters and combinations that are not part of the limited character sets that physical keyboards and keyboards are primarily designed for. It's not that there aren't ways of facilitating input - virtual keyboard layouts (keyboard drivers that can be designed like and shared, like Keyman, and onscreen keyboards) have been with us for years, and there are other input systems (voice recognition / speech-to-text being one). The problem is lack of standard arrangements and systems for many languages. Or perhaps in the matter of input systems, the old wag, "the nice thing about standards is there are so many to choose from," applies.

The result, arguably, may be a drag on widespread use of extended Latin characters, and as a consequence of popular use on digital devices of languages whose orthographies include them. Or a choice to ASCIIfy text (using only basic Latin), as has been the case with Hausa on international radio websites. Or even confusion based on continued use of outdated 8-bit font + keyboard driver systems, as witnessed in at least one case with Bambara (see discussion and example).

What can the level of contributions to African language editions of Wikipedia tell us about the effect of extended Latin? This will be explored in the next post: Extended Latin & African language Wikipedias.

---

1. For example some works on forest flora which had lists of common names in major languages of the region.
2. Arguably in the case of a language written in two or three different scripts, one could have a system in each script and an accepted way to transliterate between or among them.
3. The only other prominent use I found of the term "reference alphabet" was that of the ITU for their version of ISO 646 (basically the same as ASCII): "International Reference Alphabet." The concept of reference alphabet seems to be a useful one in contexts where many languages are spoken and writing systems for them aren't yet established.
4. This approach - adopting a standard or reference alphabet for numerous languages - was taken by various African countries, for example Cameroon and Nigeria. These efforts were without doubt influenced by the process of which Niamey and the ARA were part.
5. By comparison, the Africa Alphabet had 11 modified letters and did not use diacritics. All 11 of the characters added in the Africa Alphabet were incorporated in the ARA. It is worth noting that in the range of modified letters / special characters created over the years, some are incorporated into many orthographies, others fewer, and some are rarely used if at all.

"Wogbɛ Jɛkɛ" & Ghanaian language input support

Came across mention on Twitter of the Ghanaian play "Wogbɛ Jɛkɛ - A Tale of Two Men" but with the Ga words in the title written "Wogb3 j3k3":

#Wogb3 J3k3:The Tale of Two Men is running again tonight at the ETS DRAMA STUDIO on Legon campus at exactly 7:30pm pic.twitter.com/eN3xK467Mm

— Akwesi Agyeman (@BigYinks) October 7, 2016

In fact, looking at Twitter and at the web via a Google search, one notes both this workaround and the correct spelling, as well as the ASCIIfied version, "wogbe jeke."

7 vowels and a 5 vowel keyboard

Ga, a Ga-Dangme language of southernmost Ghana, has a complex vowel system, with seven vowels distinguished in its writing system: a; e; i; o; and u; plus ɛ ("open e") and ɔ ("open o"). The latter two are used to write many other African languages such as Akan, Ewe, Mende, Bambara, and Lingala.¹ (These characters, like a number of other Latin letters, are also in the International Phonetic Alphabet.)

Many fonts include the ɛ and ɔ, however typing them is not facilitated by standard keyboards. There are keyboard layouts specially conceived for Ga (see below for a list), as well as for Akan, Ewe, and others. However, there apparently are not any keyboards to enable multilingual input - such as an Akan title included in a tweet in English. Or if there are, they are not widely used. Hence resort to "3" for "ɛ" and ")" (the right parentheses) for "ɔ."

In African Languages in a Digital Age (p. 61) I outlined several workarounds for text including extended Latin characters not supported in fonts or input systems, a summary that was a revision of something published a decade earlier.² I had not, however, noted the use of numbers or symbols among the "substitution solutions." Ade Sawyerr, who has worked with Ga input issues, mentions observing these particular substitutions - "3" and ")" - as well as others, such as "rj" for the letter "ŋ" ("eng"), which is also used in Ga.

In any event, the resort in the mid-2010s to 3's and )'s to type words in languages like Ga, Akan, and Ewe that use them is evidence of missing input options on the devices used, or inconvenience of existing options, or perhaps lack of awareness of available keyboard apps on the part of users.

Some keyboard layouts for Ga

Over the last couple of decades, and especially since the availability of keyboard utilities like Keyman and Microsoft Keyboard Layout Creator (MSKLC), there have been many keyboard layouts developed for languages such as those of Ghana that have extended Latin orthographies. A full discussion is beyond this blog post, but generally speaking, keyboards incorporating characters not on the standard computer keyboards work either through changing key assignments (such as "q" is not used in Ga, so "ŋ" is substituted for it) or via a combination or sequence of key strokes. The solution with changed keys seems to be more common on mobile device applications, whereas both approaches are found in keyboard layouts used on computers.

Kasahorow Android keyboards
menu selection

A selection of Ga keyboards:

• Online, works on-screen with keyboard or point and click; produces text for copy-paste (all 4 keyboards look to be the same)
• Free-online-writing.com Ga Virtual Keyboard
• Get2home.com Ga Keyboard
• Keyboardingonline.net Online Ga Keyboard
• Loderi.com Ga Virtual Keyboard Online
• Downloadable Ga keyboard for Windows, made with MSKLC via Kasahorow
• Ghanaian languages keyboards, which include Ga, for Android from via Kasahorow
• Ghana Keyboards, which include Dangme, via Keyman

There likely are others for Ga (and the closely related Dangme). There definitely are a number for other languages of Ghana such as Akan (or its varieties, Twi Ashanti, Twi Akuapem, and Fante), Ewe, and Dagaare.

However, more could be done to facilitate multilingual typing, so that one doesn't have to switch keyboards or keep track of key sequences to insert something like Wogbɛ Jɛkɛ in an English tweet, or say a Hausa word with a hooked letter in a text in Akan (hooked letters are not part of the Akan orthography). Could for example an extra line of keys be added to touchscreen keyboards - say on a Ghana English keyboard - with the extra characters needed for Ghanaian languages?

About "Wogbɛ Jɛkɛ"

Wogbɛ jɛkɛ is a Ga term with meanings of "we have come from far" and "our journey is still long." It is used in the title of two plays written by Chief Abdul Moomen Muslim about the historical events, beginning with "Wogbɛ Jɛkɛ: Birth of a Nation," which depicts pre-colonial history of what is now Ghana, and followed by "Wogbɛ Jɛkɛ: The Tale of Two Men," which is centered around the stories of J.B. Danquah and Kwame Nkrumah during Ghana's independence struggle.

---

1. Some Nigerian languages like Yoruba and Igbo instead use sub-dotted characters - ẹ and ọ - for these vowels.
2. Don Osborn, 2001, "The knotty problem of using African languages for e-mail and internet," Balancing Act News Update, 69.

Touchscreen keyboards for African languages?

Who's doing what with touchscreen (virtual) keyboards for African languages? Although not a big fan of them for my personal use, I've long recognized that in principle they offer the potential to bypass the obvious limitations of physical computer keyboards - with their fixed number of keys - for layouts supporting various extended Latin character sets. This is important for a number of major languages of West and Central Africa whose orthographies include extended characters.

It was a friend's recent mention of someone else's mention of developing a Hausa keyboard for Nuance Swype that has me trying to get up to speed on where virtual keyboards for touchscreens are now, with particular reference to African languages.

If your keyboard is basically an interactive picture, then it would seem to offer the potential to modify and modestly expand to accommodate different alphabets such that of Hausa (adding rather than substituting keys would be easier on a larger screen of course - tablet rather than smartphone). Has this been done for African languages (other than Arabic, which benefits in terms of computer support from its wide international usage)?

Putting the question regarding Hausa - with additional reference to Apple touchscreen keyboards - to the old Hausa charsets & keyboards forum got helpful replies from Tom Gewecke (whose Multilingual Mac blog is an excellent source of info on iOS as well as Mac for diverse languages). Seems that iPhone and iPad support for a language like Hausa that is written with extended Latin (leaving aside the question of Ajami for another post) requires apps, and then cutting and pasting to use what one produces with them into whatever one is working on (email, document).

But there's lot's more going on with touchscreen keyboards. Looking for more information on Swype (which claims support for 70+ languages), I'm coming across mention of a number of other touchscreen keyboards and character selection apps. This is a question I plan to come back to but would appreciate any information from others.