Language Memory

;; ;;

I have recently been developing some works crossing art and archive, sound and liguistics. 

Language Memory started out as a project digitizing the skolt sámi dictionary for an "art in public spaces" installation built in the foajé of the Skolt sámi museum in Neiden, north-east Norway. Basically the artwork consists of a screen and a button. When the visitors press the button a video of a local Skolt sámi representative is shown (from the Finish side of the border, since the language is no longer spoken in Norway and is currently an endangered language). He or she reads one word from the dictionary back to the visitor. Approximately 6000 words have been recorded and sorted in a database combined with dictionary data. One reader per letter of the alphabet. A lot of work went into developing computer programs and tools to make this possible, both on the field-recording side and the presentation side.

Currently I am working on an internet version of this project on, but it will not be finished until later this year and will offer other configurations of the same archive.

There is also great interest from the linguistic community in this project because of the tools developed may be used in other settings, with other endangered languages. The material collected will also be stored as a part of the Dobes-archive ( in Holland.

Next week I would like to invite everyone within reasonable distance to two events:

27. September: Word/Play - the Sound of Dead Languages, lecture at KHIB, Bergen, Norway

30. September: Phonophani, Full-Pull Festival, Malmö, Sweden

The Sound of Dead Languages by phonophani A mini-LP with a collection of various experiments based on audio clips of dead or dying languages. What is a language? What do we loose when a language dies? If not any meaning, then at least the unique sound. Each track tries to extract harmony, rhythm and timbre from recordings made of the last speakers of various languges. The set is released under a Creative Commons License.

Some more Phonophani dates for March:

5. March: HKS, Bergen (music for Len Lye film).
11. March: Knipsu, Bergen (presentation/performance "the sound of dead languages")
23. March: Blå, Oslo (with Pierre Bastien)
24. March: Borealis/Landmark, Bergen (with Pierre Bastien)
27. March: Presences Electronique / GRM, Paris
31. March: Athens (with Pierre Bastien)

See you there!

The magical properties of an archive

Dismembering the semiotic, communicative from the phonetic, lexical aspect of language opens up a possibility for magical correspondences. The onomatopoetic, the alphabetic, the mimetic. The mysterious shapes of individual letters, the picture puzzle of the word. Language becomes an archive of non-sensuous similarities ready for the reader that connects the dots. The reader then becomes a bearer, a medium for the magical aspect of a shadow language.

Language and time

The beginning of the Skolt Saami Language Memory Project speeds up history, that is, the inevitable entropy of a endangered language. The end of the project slows it down again to the point of exhibiting a frozen distribution of a language in time.

Usually one considers language to be distributed in space, by its agents in their geographical area. One can also consider the history of a language - its development and transformation. But seldom, if ever, do one witness a language distributed in time. One word at a time – throughout the months, or years – the installation in the East Sámi Museum will parse through the dictionary depending on the amount of visitors passing by.

(One side effect of this distribution will be the disassociation of the language from the normal identity discourse of indigenous people. Legally, to be counted as Saami you have to document that your parents, grandparents or great-grandparents spoke a Saami language. Language is the principal marker of your identity, which may increase cultural isolation. In this case the language will be given to all – putting into question this identity marker).

A destructive perspective.

Today: the first tests of the recording and archiving system. Everything has to work perfectly before we take it into the field in a couple of months. The informants will be filmed in their home surroundings looking into the camera. The words of the dictionary will appear on the screen before them and they will read them aloud one by one. Each word will be stored as a separate video file on the computer.

A dictionary is in essence artificial. Only some rare kinds of poetry can bring life into a list of words starting with the same letter, and even then it is seldom systematically alphabetical in its construction. The alphabet and written language in general has a stench of death about it.

The choice of the dictionary as the image of language is the complete opposite of language-as-life. If languages are organic and alive by nature then the Language Memory Project would seem to spell out a death sentence for the Skolt Saami language.

To make matters even worse I am asking 30 representatives of the language in question to become dictionary robots reading aloud only the individual words – the atoms of their living language – in a room with no listeners. On the Finish side this will involve about 10% of the Skolt Saami community – on the Russian side it will involve 100%. In effect they will be atomizing their own language into a list of dead, alphabetized items.

All the while I will be there silently filming the spectacle – this burning funeral pyre of a self-destructing language.

Phonetic alchemy

The metaphor of dying and living languages is based on a dated romantic conception about languages being organic in structure. Thus there is a need to revisit the dialectic of the death and life of languages to view the language-image from a fresh angle. The philosopher Walter Benjamin writes in the introduction to his essay on Goethe’s Elective Affinities:

– The history of works prepares for their critique, and thus historical distance increases their power. If, to use a simile, one views the growing work as a burning funeral pyre, then the commentator stands before it like a chemist, the critic like an alchemist. Whereas, for the former, wood and ash remain the sole objects of his analysis, for the latter only the flame itself preserves an enigma: that of what is alive. Thus, the critic inquires into the truth, whose living flame continues to burn over the heavy logs of what is past and the light ashes of what has been experienced.

Replacing the concept of “work” with “language” in this quote one can perhaps glimpse a more complex dialectic at play. According to Benjamin’s idea of a critique and the work of the critic it would follow that the critic does the exact same historic and linguistic analysis as the history scientist, or “commentator”. But still their aim and result is vastly different. The critic use the detailed analysis as a means to destroy and dismember the wholeness of the work. This is made more potent by the history that has gone before. The more obscure and forgotten the work, the better suited it is for a philosophical and artistic critique. The resulting destruction-through-analysis is comparable to an archive of language: a dictionary or database of language samples, each analysed into every last miniscule phoneme. The archive kills the living language in order to preserve it, but at the same moment creates its potential alchemical transformation into new life.


– The amount of words in Skolt (as in all other living languages) is infinite, explains Michael Riessler, head of the Kola Saami Documentation Project in our first email conversation.

– And besides this every speaker of Skolt has its own stock of words in her or his mind. If you restrict to the word stock found in the existing dictionaries of Skolt (and ignoring that not all words found in the dictionaries are representative of all the single speakers' Skolt Saami language) you end up with approx. 10000 recorded words multiplied to more then 30 letters of the alphabet. Every linguist will envy you for such a collection of recorded words!

The facts are not very uplifting: There are 4 Saami languages spoken in the Kola Region (including northern Finland): Skolt, Akkala, Kildin, and Ter. Akkala is now extinct. The last speaker of Akkala passed away in 2003. Ter Saami in the Murmansk region has about 30 speakers, all age 50 and above. The Kildin Saami has about 300 active speakers. Likewise the Skolt Saami has around 300 speakers on the Finish side of the border and only a handful of old people left on the Russian side (speaking the special Russian dialect of Skolt).

But Michaels comments make me rethink my melancholic impression of Skolt Saami as a dying language. Infinity is a powerful concept to bring into any reflection. If every speakers vocabulary is potentially infinite or to be regarded as a “part” of infinity (that in itself would be infinite), then a dying language is not ceasing to exist by slowly shrinking in size as one would expect (due to forgetfulness, language shift or some other kind of deterioration of the collective memory). It is still present and alive in its vibrant infinity even with only one speaker left on earth (or maybe two? Does not a language need a listener? Or maybe it is sufficient with only one subject speaking to him- or herself? I guess that would be the perfect communication: The last speaker of a dead language muttering to himself).

I feel my eyes drying up. I am lost in a desert of broken letters. Literarily.

Struck by a sudden premonition I see my next two weeks before me: working day and night proofreading a dictionary that translates between two languages, neither of which I understand a single word.

Today has been a technical research day. How to digitise a dictionary. How to wield the computing power to my needs. How to teach a blind computer to read.

The first goal: to build a database of the Skolt Saami language. A dying language in the arctic regions of northern Norway, Finland and Russia.

The next goal: To travel to northern Finland and collect samples of all the words of the Skolt Saami language. To film one informant per letter of the alphabet reading all the words.

The final goal: To build an art installation for the new Skolt Saami museum in Neiden, Norway - an exhibition of the totality of a dying language. At the push of a button, one by one of the words will be given to the visitors. One word each, given as a task - for the visitor to take responsibility for and remember for the future.

The beauty of totality.

By now I probably own the largest library of Skolt Saami to Finish language dictionaries in the world. Except for a few Skolt - German ones from the 1800 century that I found visiting the Humbolt University Library in Berlin last month, I have gathered all I could find.

The sum total is 4 books and 1 bad photocopy from the 80s.

They all vary greatly in size and quality, and I have tested my way through them all in the hopes of finding a candidate for scanning and optical character recogniction.

Finally today, a breakthrough. Mosnikoffs and Sammallahtis dictionary from 88 seems to have all the necessary ingredients: the copy is in strong black and white ink, the “c” does not look like an “e” (who would have thought that this would be the greatest of challenges for the digitizing community?) and all the special letters of the Skolt Sami language are possible to separate from each other. The ? from the d, the ? from the k, the ? from the ?, the š and ž, and å and â, not to mention the õ ö ? and ?. (Even your browser probably do not support these characters...)

I am starting to get a close relationship with these characters. Their corresponding sounds roll silently in my mouth like liquorice while I stare at the enlarged scans. Teaching the computer to understand all of them takes patience, but gives a rare glimpse into the microscopic world of the letter. The shapes of the characters are blown up and supersized until I see every molecule of ink filling up the topography of the paper.