1. The current status of Icelandic

Iceland is one of the smallest independent language communities in the Western world. Icelandic, which is the only official language of Iceland apart from Icelandic Sign Language, is only spoken by around 340,000 people. However, Icelandic has until very recently been the first language of virtually all inhabitants. The language community is very homogeneous, and regional dialectal variation is negligible. Icelandic has changed less than related languages, although the changes are much more extensive than we often pretend, and than most Icelanders think, and Icelanders are known for their language purism.

A few years ago, we could maintain that Icelandic was used – and was in fact the only language used – in all spheres of society; in government and administration; in education; in business and commerce; in the mass media; in cultural life; on the Internet; and in general face-to-face communication. But does this mean that the language is safe?

Several attempts have been made at estimating the vitality of the world’s languages, and a number of vitality scales have been proposed. Among the best known are UNESCO’s Language Vitality Scale which was first published in 2003, and the EGIDS, or Expanded Graded Intergenerational Disruption Scale, which was put forward in 2010 as a revision of Fishman’s GIDS Scale from 1991. In the UNESCO scale, Icelandic fulfils the criteria for reaching the highest level in all the six major evaluative factors that the scale is based on. In the EGIDS scale, Icelandic is on the next highest level. Thus, Icelandic should be on the safe side, according to both of these scales.

However, there may be some dark clouds on the horizon. Icelandic has gone through dramatic ecological changes in the last decade or so. These changes, which are both sociological and technological, have greatly increased the external pressure on the language. Let me give a short overview of the main changes that are causing this pressure.

First the sociological changes. We have had an explosion in tourism. The number of foreign tourists has increased by around 30% per year in the last five years or so. This has greatly affected our language environment. In downtown Reykjavík, for instance, everything is directed towards tourists. Menus outside restaurants are often only in English, and the same goes for all kinds of advertisements and signs. A number of cultural events, such as concerts and theatre performances, are also introduced and performed in English to attract tourists.

The number of foreign workers has grown considerably and it has recently been predicted that people of foreign origin will amount to 15% of the population in 2030. A high proportion of these people work in restaurants or as shop assistants, where they have to communicate with customers, and natives often complain that they cannot use the country’s official language in shops or restaurants because none of the staff understands it.

English is also increasingly being used in higher education – more and more university courses are taught in English. Up to now, the main reason has been the desire or need to accommodate foreign students and their number is constantly growing. Now, the number of foreigners among university professors is expanding rapidly and most of them cannot teach their courses in Icelandic. The pressure of writing academic papers in English is also increasing.

An important factor in this respect is the ongoing globalization and increasingly globalized attitudes. Young people in Iceland, and probably in most Western countries at least, think more globally than their parents did. They see the whole world as their home, their stage, their playground, their market – they want to travel abroad, study abroad, work abroad, live abroad, and then eventually return to their native country – or not.

Now to the technology. It's only a decade since the advent of smartphones, but smartphones and tablets have revolutionized the daily lives of people, especially children and teenagers who are now online 24/7, so to speak. They are directly connected to the digital world which is for the most part in English. This is perhaps the greatest cause of changes in the language environment in the history of Icelandic, and hence a potential indirect cause of changes to the language.

Research has shown that the majority of children and young people no longer watch old-fashioned linear TV but watch material from service and content providers like Netflix and YouTube instead. This is important since all programs on Icelandic TV are in Icelandic – either dubbed, as all programs intended for young children, or with Icelandic subtitles. Netflix and YouTube, on the other hand, offer very limited material in Icelandic. It is a common sight in shops and restaurants to see young children, even in carts, holding smartphones which have been handed to them in order for them to keep quiet and watch some stuff on YouTube – almost certainly in English – while their parents are shopping or enjoying a meal.

Computer games which are especially played by young people have also changed. They are now more and more interactive, which means that players are not only reacting to actions in the game, as used to be the case, but communicating – either with the game program itself, or with other players. Since these players may be spread around the globe, the language of communication is often English.

The technological change that might have the most far-reaching consequences for Icelandic is the imminent explosion in the use of voice control. Most of us have heard of and even tried to use Siri in iPhone or some of the new digital assistants – Amazon Alexa, Google Home, or Microsoft Cortana. YouTube is full of funny videos which show people trying to use voice control – unsuccessfully, because they have Scottish accent, because they have just come from the dentist and their pronunciation is slurred, and so on. However, this ceases to be funny when we realize that tomorrow, this will be our reality – everything around us will be voice controlled. Our cell phone, our computer, our car, our household appliances – you name it. What will that entail for speakers of a language like Icelandic, which is currently unusable in this new technology?

2. Possible consequences of external pressure

It remains to be seen how or whether these radical sociological and technological changes will affect the Icelandic language. However, we can imagine at least four different types of effects these changes might have.

It is likely that direct influence from English will increase, and we are already witnessing some of this. I’m talking about changes in vocabulary – more and more English loanwords, some of them adapted to Icelandic phonological and morphological rules, but others not. The inflectional system might also be affected in different ways – weakened by loss of some inflections, and changed by the introduction of English inflections such as plural –s-ending. The syntax might also be affected, especially word order and fixed expressions; and the phonology might be affected, for instance by relaxation of some phonotactic constraints, by weakening of the Icelandic stress rules, etc.

Icelandic might also lose domains of language use. I have already mentioned that Icelandic cannot be used for voice control. Ten years ago, English was becoming the working language in many Icelandic companies. This development had a backlash in the financial crash in 2008, but I think it’s returning. English is also the main language used in the tourist industry which has become the most important industry in Iceland in the past decade. Furthermore, English is becoming more and more dominant in higher education, as I mentioned previously.

Other effects might be even more serious. We know that children in the language acquisition period – say up to six to nine years of age – need a certain amount of language input to acquire the language of the community and build their own solid internal grammar. With English dominating their digital world – on the Internet, in smartphones, in computer games, on YouTube, etc. – chances are that the children’s exposure to spoken Icelandic will be considerably lessened, perhaps below the necessary threshold for successful first language acquisition. In addition, their parents might be preoccupied with their smartphones instead of talking and responding to their children. Theoretically at least, this might result in a group of children – in the worst case, a whole generation – growing up without mother tongue competence in Icelandic. This might lead to greater variation in the language and eventually to degradation of the grammar system.

Finally, Icelandic might lose prestige, especially among young people who know from the outset that Icelandic is useless abroad. I think they can live with that but what if it turns out that they cannot use Icelandic at home either? If Icelandic will be excluded from the digital world of the future, where these people want to find new and exciting job opportunities and entertainment, the effects on their attitudes towards Icelandic could be very negative. They might find the language old-fashioned and incomplete, and this might pave the way to the death of the language. It is well known that speakers’ attitudes are one of the key factors in determining the vitality of the language. If speakers generally foster negative attitude towards their language, it is doomed. The attitudes of young speakers are particularly important in this respect.

So, is there reason to worry? It is often pointed out that Icelandic has been in close contact with other languages in the past without suffering severe damage. Let me just mention two examples.

When the famous Danish linguist Rasmus Christian Rask came to Iceland in 1813 he came to the conclusion that contact with Danish had already had great effects on Icelandic. He predicted that no one in Reykjavík would be speaking Icelandic in one hundred years’ time, in the early 20^th century, and no one at all in the whole country after three hundred years, that is, about one hundred years from now. It remains to be seen whether the second part of Rask’s prediction comes true, but the first part certainly didn’t – Icelandic is still spoken in Reykjavík, more than hundred years after it ought to have vanished from the city according to Rask.

In the 1960s, the American naval base in Keflavík had its own TV station which was the only TV station in Iceland for a few years – the State Television was only founded in 1966. People in the neighbourhood, including Reykjavík, could watch its broadcast, which was of course in English. Many people were worried that watching this station would seriously affect our culture and especially harm the Icelandic language – but it didn’t. However, it must be kept in mind that the effects of this station only lasted for a few years, and after the Icelandic TV started broadcasting, the broadcast domain of the American station was limited to the naval base.

Even though Icelandic has survived close contact with Danish and English in the past, that doesn’t mean that we shouldn’t be concerned now. There are at least three reasons for that. The first reason is the sheer quantity of the English input we receive. English is all around us, both in the digital world and in the material world, all the time. Thus, the pressure from English is more voluminous and more comprehensive than anytime before in our history.

The second reason is that the recipients are younger than before through smartphones, tablets, video games, etc. The influence from Danish in the 19^th century, for instance, only affected a limited group of adult speakers. Effects of close language contact on children in the language acquisition period might be much more serious.

The third reason is the nature of the contact. The use of English is becoming more and more interactive instead of the passive reception of the old days. Research has shown that receptive input, like watching TV, does not provide enough stimulus for successful first language acquisition. When children start conversing in English, either directly or through computer games, for instance, the effects of the language contact on their mother tongue might be more deeply rooted.

So there may, in fact, be reasons to worry, and in the past few years, the prospective future of the Icelandic language has become the subject of much discussion in Icelandic media. Three years ago, a speech pathologist working with children, Linda Björk Markúsardóttir, wrote an article in the newspaper Fréttablaðið, where she said:

“One of my tasks in my job as a speech pathologist is assessing the vocabulary and language skills of children and teenagers. I increasingly meet Icelandic children who only have a rather superficial knowledge of Icelandic. When I show them pictures of common things and ask them to tell me what these things are called, I often get answers like: “I know what this is, see, I just don‘t know the Icelandic word.” If they are offered the option to give the English word instead, they answer spontaneously. How can this be? How is it possible that children and teenagers who have Icelandic parents, are born and raised here and have attended Icelandic schools from the beginning have not acquired basic Icelandic vocabulary? [T]he answer is simple: Computers and technology. [C]hildren and teenagers live in the world of digital communication which almost entirely occurs in English.”

Last year, another leading newspaper, Morgunblaðið, published a news article on children’s use of English in their conversations.

“A primary school teacher in Hafnarfjörður says that many pupils have started using English for communication at school. This applies both to teenagers and younger pupils. The teacher is not sure that Icelandic will survive current threats. He says that this started to emerge two or three years ago. He believes that the Internet is to blame. The youngsters are spending a lot of time on the Internet.”

3. The MoLiCoDiLaCo project

We hear a lot of such stories but we don‘t know how typical this situation is. Of course, people tend to notice something they find unusual and generalize from it, but for two Icelandic children speaking English among themselves, there may be two thousand, or twenty thousand, who never do such things. In order to map the current status of Icelandic and English in the Icelandic language community, we have initiated a large research project entitled “Modeling the Linguistic Consequences of Digital Language Contact”, or “MoLiCoDiLaCo” for short.

As the title indicates, we believe that the digital pressure from English that Icelandic is exposed to might have serious effects on the language. This project received a large three year Grant of Excellence, around 900,000 Euros, from the Icelandic Research Fund in 2016. Professor Sigríður Sigurjónsdóttir is the main leader of this project and I am the co-leader, but we collaborate with a number of researchers, both in Iceland, the US, the UK and Ireland, and we have a number of graduate students working on the project.

The project started officially on July 1^st, 2016. The first year was mainly used for preparatory work, such as designing questionnaires, but since July last year, we have been busy collecting data. We collaborate with the Social Science Research Institute at the University of Iceland, which provided us with a randomly selected stratified sample from the National Registry – 5,000 people divided into 11 age groups, from three years of age upwards – the oldest participant was 98 years old. These people were sent a letter asking them to participate in an online survey which took about 25-30 minutes to answer. Participants from 13 years old and upwards took the same survey, but we had to develop four different surveys for different age groups of children. Of course, the parents of the youngest children had to answer for them, but we requested that they ask the children themselves where appropriate.

In this survey, we are focusing on a number of different variables. First, we ask about speakers’ background – their age and gender, their education, their place of living, their occupation, and also whether they own a smartphone or a tablet, how old they were when they started using the Internet, etc. We also try to uncover their attitudes towards Icelandic and English – how important they think it is to speak good Icelandic and English, whether they want to have the user interface of their smartphones and computers in Icelandic or in English, whether they can envision using English for voice control, whether they want to live in Iceland in the future, etc.

Then we try to estimate the amount of language input they receive, both in Icelandic and English – how often they listen to each language, how often they speak each language, how often they read text in each language, how often they write text in each language, whether they speak English to somebody on a regular basis, how old they were when they started learning English at school, whether they knew any English before they started learning it at school, etc.

The main part of the survey is devoted to a number of linguistic variables. We try to estimate speakers’ vocabulary by giving them lists of Icelandic and English words of different frequency and ask them to check the words they understand and can use. Finally, participants are asked to judge a number of Icelandic sentences on a five point Likert scale and say whether they find them completely natural, rather natural, neither natural nor unnatural, rather unnatural, or completely unnatural.

These sentences are of different types. We tried to select sentence types where we suspect there might be some influence from English. We also have sentences where we know there is variation, in order to see whether English input might increase the variation. Furthermore, we have sentences with well-known “grammatical errors” which have been frowned upon in schools for a long time, such as the well-known “dative sickness” and the “new impersonal construction” or “new passive”. We also have a few English sentences of types where Icelanders often make errors by transferring Icelandic syntax to English. We wanted to see whether the frequency of such errors would be correlated with the amount of English input speakers receive.

The online survey has now been closed, with a turnout of around 41% for adults and 50% for children, and we are working on the next phase in data collection – interviews. We have selected a group of 400 people from the participants in the online survey, according to certain criteria. We are conducting deep interviews with these people – two times two hours interviews with the adults, and 3 times one hour interviews with the children. In these interviews, we are mainly investigating the same variables as in the online survey, but in much more detail, and using different questions and methods. The interviews with the adults are almost finished, and the interviews with the children are underway and we hope to finish them by the end of this year.

In addition to the online survey and the deep interviews, we use focus groups and interviews with primary school teachers and pupils. We also plan to open our online survey to anyone interested and use social media to distribute it. We hope to get up to 30,000 people to participate in the survey. This group will of course be kept separate from the original participants in the survey which were selected from the National Registry according to certain criteria, as I mentioned.

We have only just started processing and interpreting the data from our online survey but I can briefly mention a few of our first results. Almost 20% of three to five year old children use smartphones or computers more than one hour per day. The number increases with age, so almost 40% of six to seven year old children use smartphones or computers more than one hour per day, and more than 60% of eight to nine year old children. 10-12% of teenagers, the groups 13-15 and 16-20 years old, use smartphones or computers more than eight hours per day.

Let’s turn to speakers’ attitudes. Almost 70% of 13-15 year old speakers either agree or strongly agree that they could envision using English for voice control. On the other hand, almost 80% of speakers 60 years old or more either disagree or strongly disagree with this assertion. We get a similar pattern for the user interface. Almost 50% of 13-15 year old speakers either agree or strongly agree that they prefer an English interface on their computers and smartphones. Conversely, more than 70% of speakers 60 years old or more either disagree or strongly disagree with this assertion.

Finally, let me mention a sentence type where we have had some syntactic variation for a long time, but where we think that an increase in English input, accompanied by less Icelandic input, might speed up an ongoing change. This is the use of the indicative in subordinate clauses where the subjunctive has traditionally been used. Younger speakers accept the indicative in such sentences to a much greater extent than older speakers. The reverse tendency, that is, to use the subjunctive where indicative is the norm, also exists and is more prominent among younger speakers.

One of our graduate students, Elín Þórsdóttir, looked closer at this in her master’s thesis and she found out that for speakers from 16-20 years old, there is correlation between their exposure to English and their “untraditional” use of moods. Those speakers who report most exposure to English also show greatest variation in the use of moods.

4. Icelandic language technology

I don’t have time to mention more examples from our research, but we will of course be publishing our results in journals and books in the near future. Let’s turn again to the status and future of Icelandic in the digital world, because I believe that it is a matter of utmost importance to make sure that Icelandic enters the digital realm.

In April 2017, Associated Press distributed an article entitled “Icelandic language at risk; robots, computers can’t grasp it”. This article was published under different titles in the media in at least 30 countries, among them both The New York Times and The Washington Post. Last February, The Guardian published an article entitled “Icelandic language battles threat of 'digital extinction'”. Both articles emphasize the threats that the digital revolution poses for a small language like Icelandic.

It has recently been claimed that at least 95% of the world’s seven thousand languages or so are awaiting so-called “digital death”. This means that these languages don’t keep up with the rapid developments in language and speech technology and will slowly disappear from the digital realm – or fail to enter that realm altogether. They will not be used on the Internet or in computer games, and language technology tools and applications such as speech recognizers, translation systems and so on will not be developed for them.

Of course, this claim is only a prediction, even though it is supported by empirical evidence, and things might turn out differently if the necessary measures are taken. It must also be emphasized that it is not clear what digital death entails – does it necessarily lead eventually to death in the real world?

Iceland was an active participant in the European META-NET project from 2011-2013. One of the main deliveries of META-NET were the so-called Language White Papers describing the situation of 30 European languages with respect to available language technology support. The main conclusion of the White Papers was that most European languages are unlikely to survive in the digital age – at least 21 official European languages are in danger of digital extinction.

The paper on Icelandic highlighted the alarming status of Icelandic language technology. Icelandic was among four languages which received the lowest score, “support is weak or non-existent” in all four areas that were assessed – automatic translation, speech interaction, text analysis and the availability of language resources.

The White Paper received considerable attention in Icelandic media and its results were discussed in the Icelandic Parliament. Two years ago, the Minister of Education, Science and Culture appointed a special language technology steering group. The group commissioned three language technology experts to work out a detailed five-year Project Plan for Icelandic language technology. This plan proposes that emphasis is laid on five core tasks: Speech recognition, speech synthesis, machine translation, language and style checking, and language resources. In addition, it is proposed that a competitive research and development fund will be established.

In the policy statement of the new government that came to power last November, it is explicitly stated that this plan will be implemented and financed. The total cost of the Language Technology Project Plan is estimated to be approximately 18 million Euros. On the budget law for 2018, one fifth of that sum, approximately 3.6 million Euros, is allocated to language technology, and a similar amount on the budget proposal for 2019.

The Project Plan proposes that the self-owned foundation Almannarómur will be commissioned to implement the Language Technology Project. The members of Almannarómur are universities and research institutions, IT companies, financial institutions, insurance companies, energy companies, companies in the travel industry, and organizations of disabled people.

An agreement between the Ministry of Education, Science and Culture and Almannarómur was signed on August 20th. The role of Almannarómur will be to conduct the five above-mentioned core tasks. The Research and Development Fund, on the other hand, will be under the auspices of the Icelandic Centre for Research. We hope that the first calls can be announced later this fall.

5. Conclusion – the survival of Icelandic

I have now come to the end of my talk but let me finish by emphasizing that Icelandic, like any other language, is not only a scientific and cultural treasure which needs preserving. It is essentially a social phenomenon – by far our most important tool of communication with other people. Therefore, the Icelandic language norm, which was defined in the late 19^th and early 20^th centuries, needs to be relaxed somehow. The language must not stagnate, but has to be vivid and adapt to the needs of the society at any given time. It must tolerate variation in pronunciation, inflections and syntax, and that new words enter the language and old words acquire new meaning. It must not become the private property of certain groups of people, and the language and variation in its use must not be used to discriminate people or categorize them in any way.

Last but not least, language is the most important outlet for our feelings – love and joy, hatred and anger, sorrow and sadness, hopes and desires – but also our tool for creative and innovative thinking. A language that we acquire in childhood, our mother tongue, is an integral part of ourselves, our private property but at the same time public property belonging to the whole language community, and in a sense even to the whole mankind. This sounds like a paradox – and it is a paradox. It is by no means a simple matter to treat the language in such a way that all its different roles are respected. That is, however, what we must strive to do. Despite technical advances and globalization, our mother tongue – any mother tongue – is still invaluable for us all.

I believe there are two main prerequisites for the survival of Icelandic, and the same is probably valid for many other small languages. First, we must ensure that our language is usable in the digital world in order to spur positive attitudes towards it among its speakers. Second, we must ensure that our children receive enough relevant input in our language by talking to them and urging them to use the language productively.

Icelandic is not dying – it can and must be saved, but we need a joint effort. As I have mentioned, the government has decided to invest considerably in Icelandic language technology, in order to make Icelandic digitally viable within the next five years. Now it’s up to us all to raise awareness among parents and other adults about the importance of the conversation for language acquisition. The future of Icelandic depends on our children’s possibilities to use the language in all domains – and on their willingness to do so.