Scotland and language technology

robot photoI spent Thursday afternoon in Copenhagen where the Danish Language Council was launching a report about language technologies and how to make them available in Danish.

The main point was that when companies decide which languages to launch their products in, they compare the size of the market with the difficulty of implementing it. So small languages like Danish will get ignored unless plenty of resources are accessible easily and cheaply.

They therefore recommend that the Danish state should spend money on creating and collating a lot of languages resources (such as dictionaries, corpora and annotated speech recordings) and making them available for free. Furthermore they suggested that Danish universities should create courses in Danish language technology, i.e., not just in computational linguistics, but with a strong Danish language component.

I cannot help thinking that Scotland should be doing the same. I think many people think that only Gaelic is a problem because Scotland is English-speaking, but as we all know, most spell checkers hate the good Scots word outwith, and most language speech products (such as Siri and Alexa) struggle with standard Scottish English:

I believe Scotland should be doing the same as Denmark. The report suggests spending between £1m and £2m a year on this, but it might be dearer in Scotland because there are three languages (Scottish English, Scots and Gaelic) instead of just one, and because more resources are already available in Denmark.

Leave a Reply

Your email address will not be published. Required fields are marked *