The Welsh language is a stand-out in its ability to survive - there are more Welsh speakers than there are speakers of the all other Celtic languages put together. As part of a continued effort to support the Welsh language, support for it in IT must be an important factor.
Most technologies (programming languages, configuration, names of things) are based on English, and it is unrealistic to think that this could be otherwise, but there are many places the Welsh language must be supported:
Welsh language websites - requires welsh speaking developers.
Welsh governmental and private corporations support the Welsh language, and hence put time and money into translating content at least, but these services are invariably hosted outside Wales. There is no reason for this.
A web developer with a strong desire to build a website supporting Welsh, who for patriotic reasons wants to host the website in Wales would find it difficult to do
For developments in Wales service Welsh speaking communities and businesses, it makes sense for Welsh Language to be a viable language used by those involved in IT, and providing these ubiquitous cloud services. Without doing this, the Welsh language in not participating in one of the most vibrant sectors: the ground is being surrendered to English without a murmur.
Two quotes from The Welsh Language Technology Action Plan
"To summarise: we want to see the Welsh language at the heart of innovation in digital technology. We want to be in a situation where Welsh is offered proactively, and in which families, organisations and individuals can use Welsh in an increasing number of digital contexts—be they monolingual or multilingual—without having to request it."
"To that end, the Digital Competence Framework5 will continue to offer tools and resources in Welsh and develop ways to facilitate creation of more Welsh-language content and the use of every kind of technology through the medium of Welsh"
This is a short document (30 pages) which attempts to address a monstrously large field, and is unlikely to be useful. If anything the document reads like one written by a committee that had to produce something, but gets bogged down in details of one or two ideas while woefully missing the broader picture.
Thus we get objectives like “develop a stemmer”:
“A stemmer cuts the end off words to reveal the root. For example, the Welsh language verb datblygu (develop): ‘datblygodd’, ‘datblygais’, ‘datblygiad’, ‘datblygiadau’ would be cut to the stem, which is ‘datblyg-’. Additional work is needed to deal with irregular verbs. While a lemmatizer deals with syntax, the stemmer reveals the meaning of words in sentences, the semantics. It is a useful tool in the development of machine learning and artificial understanding in Welsh.”
Fair enough, but this is a fiddling small detail that has no place in a strategy document.
Consider how speakers of other languages - let’s take French as an example - are affected.
There are French developers writing code, developing apps, making websites for the French speaking markets; French App developers writing apps for Apple iPhones in French.
There are many links from this page to other pages in French, but the critical documentation without which it is impossible to develop an app is all in English. Furthermore, the API (application program interface) which all iPhone developers must use is based on English, so App developers
regardless of the language they speak have to use functions like:
To an English speaker, the name and parameters of this function at least hint at what it is for; French speakers simply have to learn it.
Similarly with widely used programming languages such as Javascript, PHP, Ruby…..(this could be a very long list) the documentation is mostly in English, and the language itself is based on English.
Here is a link to the French wikipedia page describing the very widely used Javascript programming language, all written in French:
https://fr.wikipedia.org/wiki/JavaScript
However consider this example - from that wikipedia page - of Javascript written by a French developer:
The developer has used french names for his variables which he can define himself (monObjet, methode..) but the keywords which are part of the javascript language shown in green and red here are derived from English.
So the situation here is that in France there are French speaking developers, working in French offices, eating lunch at French Cafes, but having to have at least some familiarity with the English basis of the programming languages they use.
We could imagine a Welsh development company employing Welsh developers, speaking Welsh and engaged in Welsh culture. This is a vision that is certainly consistent with the Million Welsh speakers by 2050. So how do we achieve this?
A Welsh company employing Welsh speaking people would no doubt see the Welsh speaking community as a target market. This of course is a small market, but the same developers are of course able to develop code for the English speaking markets. This gives them an edge as companies outside Wales are unlikely to compete for bi-lingual or Welsh language applications. (A growing market of course as there will be 1 million speakers by 2050).
Developers in Wales could also hook up with many other organisations working on exciting projects - for example Bangor University has a centre for leading edge research into 5G.
So what is needed is to encourage a community of developers and IT users in Wales. Servers, cloud computing, micro-services are all now fundamental building blocks for this on which other applications - be they simple websites to advanced Artificial Intelligence applications - are built.
Investment in data centres and hosting in Wales is a strong move in this direction. Furthermore, the Welsh government should be very possessive about keeping such developments in Wales: any Welsh language website that is hosted on a server in California (because it is cheap and easy to do) is revenue and skills lost to Wales.
For Wales to have a successful data centre industry, we must ask why Wales is a suitable place for companies, charities, government agencies to store and process their data here.
The data centre services market has been growing rapidly for the past few years with a compound annual growth rate of around 13%, a trend that is likely to accelerate in coming years.
