Pedantry - Moved to http://pedantry.fistfulofeuros.net

Friday, August 29, 2003
 
Why I need an editor

Rereading my last post hours later, now that I'm back from Brussels, I hate it a lot. It flows badly and there are a lot of anaphora problems. Folks, I do usually write better than this, but I got a bit rushed to get this out this week. It's been, to some degree, pasted together out of bits written in different drafts, but without that stage where you reread the whole with a fresh brain and make them all fit together. If I had the time, I would just rewrite it now. Maybe Monday.

I have difficulty managing two big writing projects at the same time, and unfortunately IWT-Vlaanderen got my best work this week.

Oh, and David Weman over at Europundit is promising to "to demolish all [my] arguments." He doesn't know that I didn't originally start exploring this set of ideas for language policy, I developed them as a critique of representative democracy and private property for a science fiction novel that I was, once upon a time, planning to write. He hasn't even seen the really radical parts yet. If he's really interested, I'll explain why we ought to abolish elections and what I think we ought to replace them with.
 

Thursday, August 28, 2003
 
Language, Culture and Reality

Finally, my long-delayed fourth post on language rights. Click appropriately for Part 1, Part 2, and Part 3.

I'm not entirely happy with it, and may make some changes later on. It seems a bit diffuse to me. Well, I hope y'all aren't reading the blogs for polished scholarship. But, I decided that since I'm so delayed putting this out, I didn't want to wait until Monday to publish it. I have a rather busy weekend ahead, so there will probably not be much new here until next week.

I didn't quite cover everything I intended to say, but this post was already growing quite long.

Let me also say thanks to Kerim at Keywords for linking in. He also has some good points about this series. I am hoping to get together a post collecting and responding to the response I've gotten in the blogosphere.

But anyway, on with the show...




Reality is that which, when you stop believing in it, doesn't go away.

-- Philip K. Dick

Our language is dying because reality no longer takes place in Occitan. And what purpose do our words serve, when there is no longer any meaning with which to fill them?

-- Péire Pessamessa, quoted in Langue et Société by Jacques Leclerc 1

There is a notion, widespread enough outside of linguistics but almost extinct within it, that claims that language has a determining influence on how we think. Lev Vygotsky, who I discussed in my last post, was one of the people who believed this. It's a really tempting notion and in its own way, it could explain quite a lot. Why do those pesky Arabs, Mexicans, Indians, English, whoever, do whatever it is they do that seems so incomprehensible? Blame their language.

This theory is often called the Sapir-Whorf hypothesis, but I've always preferred to call it "vulgar Whorfism." I am not a Whorf scholar, but I have known a couple who quite earnestly claim that what most people take to be the Sapir-Whorf hypothesis isn't what Whorf said at all. It's just a contagious idea that people have read into Whorf. This may be true - it's happened to other people - so I stick to "vulgar Whorfism" just in case.

Vulgar Whorfism is a really very tempting notion, and it's always reinforced by the same example: The 20 [30, 50, 100 or more] words the Eskimos have for snow.

First of all, let me be politically incorrect for a moment. The word "eskimo" does not derive from the Abenaki word for "bad breath", something even many Eskimos believe. No one's quite sure where the word comes from. It came into English via the French esquimaux, but where the French got it from is a bit of a mystery. The two major theories are that they got it from the Spanish esquimal or from the Montagnais word for snowshoes. Still, folks like to be called what they like to be called, and Eastern Canadian Eskimos like to be called Inuit. In an etymological sense, this word is even more disturbing than "eskimo." It is the Inuktitut plural noun for "human", so all us non-Inuk are what?

I actually spent four years of my childhood in the city of Iqaluit (back then still called Frobisher Bay). My father was the shop teacher at the high school, and my mother taught first grade. Paul Okalik, the Premier of Nunavut, was one of my father's students.

I've travelled a lot in life, but Iqaluit is easily the most exotic and out of the way place I ever lived, and I have a residual attachment to Inuit issues. Back when I was a student in Montreal, and I still entertained the idea of becoming Canada's first NDP prime minister, I decided that having an aboriginal Canadian language in addition to English, French and Chinese (which I was studying at the time) couldn't hurt me at the polls. So, I took up Inuktitut.

It's an interesting language in a lot of ways. It has a lot of really interesting linguistic features. The one that I always found the most interesting is the "fourth person." Inuktitut has split the third person - he, she, it, them - into two different pronouns and morphological categories (actually, the pronouns are the morphological elements in Inuktitut), one for the main person in a narrative, and one for any other person who intervenes in the narrative. This notion has always brought to my mind Samuel Delaney's science fiction novels, where he takes many languages issues far more seriously than most other SF authors. I've long thought there might be an interesting bit of science fiction in a story about the ways language can have syntactic and morphological elements that specifically serve narrative functions.

For those of you who have never seen or heard Inuktitut, here's a little snippet of printed text lifted from the Nunavut government website:


The above reads Nunavut Guvamangit - I'm surprised that I only had to look up two characters - "Government of Nunavut." It looks like something you'd expect to see on an alien starship. Back when I was kid, I thought that was way cool. You would hardly think that this syllabary derives from Pitman shorthand. This same syllabic writing system, modified to some degree, is used for a dozen or so Canadian aboriginal languages. Most dialects of Cree are written using it, as is Ojibwa, Chippewyan, both Slavey dialects, Athapascan, Carrier and a number of smaller languages. It is the only writing system unique to Canada.

I've forgotten a lot, but there is one thing I haven't forgotten: There is only one good word for snow in Inuktitut. It is "aput" in its canonical form - or at least it is in semi-standard Eastern Arctic Canadian Inuktitut. If you live in Greenland or anywhere west of Coppermine, your mileage may vary.

The thing is, even if Inuktitut had a zillion words for snow, it wouldn't have proven anything per se. You see, vulgar Whorfism is almost true. The difference is so small, and yet so important. It is an idea turned upside-down and I want to set it right-side up. The structure of our language doesn't fixes the content of our thoughts, our thoughts fix the structure of our language.

Language is not merely a protocol of communication. It isn't simply a code. It is the reflection of our mental life. Vygotsky explains how our interactions with others is a way in which we use the structures around us to expand our powers in every way, taking them into ourselves. And, language is the foremost way in which we interact with the things that most empower us: other people. In order to make the most of the people around us, we must express our thoughts because it is almost exclusively through language that we are able to use other people's heads to think. Language can only work when it is able to express whatever we might wish to say.

Jerry Fodor is most associated these days with the idea of a "language of thought." However, he imagines it to be a symbolic repertoire and a set of combinatory rules hard-wired in some sense into the human nervous system. What I am advocating is quite different. I don't think it's useful to envision the activity of the nervous system in that way at all. I think it is useful to envision social cognition as something mediated by language, but unlike Whorf or Fodor, I don't think that linguistic mediation creates any meaningful restraint on cognition at all - in or out of the body.

Inuktitut has a single general word for "snow" because sometimes people have thoughts that concern snow in general and have a need to express them. They also have structures - sometimes independent lexemes, sometimes not - for more specific varieties of snow and specific contexts that snow might find itself in, but then so do we in English. If you live in Canada, these words have quite specific meanings, while I once had to explain to a life-long Californian what "sleet" means. People in southern California usually have little need to discuss the finer points of snow, unless they are skiers.

Inuktitut also has a single word for "snowmobile." They have no trouble manipulating this concept - nor gasoline or engine mechanics - even though it was utterly alien to them less than 50 years ago. As soon as it became necessary to talk about snowmobiles, gas and carburetors, a way of expressing those things developed, using a combination of terms borrowed from Bombardier technical manuals and invented or repurposed Inuktitut words.

The idea that language reflects the thoughts of speakers is very important. Speakers' categories of thought reflect the way they interact with the world and how you interact with the world may vary a lot depending on who you are and what context you live in. Language is not an arbitrary protocol, it is the concrete manifestation of the worldview of the culture that supports it. It is an expression of the reality of its speakers' lives.

This notion that language reflects culture without restraining it is important, because on it rests a lot of baggage. It is the principal justification for associating language with culture at all. It is also what distinguishes an advocacy of linguistic diversity from a reinvention of "separate but equal."

Culture is one of those notions that is widely used - so widely that we probably should conclude that it actually refers to something even if only in the minds of the word's users - yet so hard to pin down as to be undefinable. A variety of definitions of "culture" have been proposed over the years, but each revolves around a small set of core notions: meaning, behaviour, tradition, convention and continuity.

I want to - in a sense - abolish the noun "culture" and replace it with an adjective: "cultural." The ambiguity in the way people use the word "culture" suggests that while it probably represents some real thing for speakers - some category of phenomena that they can productively manipulate and talk about - it hides some more analytically useful idea. I suggest that what it hides is Vygotskyan tools and their cultural nature.

I want to define a culture as the set of tools - cognitive and otherwise - that people find at their disposal. I want to get rid of this notion of culture as a monolithic thing. "Cultural", instead, is an adjective used to highlight the contextually situated nature of something, particularly how it is situated within a social and historical context. That separates it from something that is merely a natural phenomena whose existence is independent of human action. Pine trees in a wild forest are not cultural, they are merely natural. A house built out of pine lumber is cultural.

When we speak of a culture as if it was a single entity, we are not designating it that way because its members form a collective. We are merely asserting that there exists some set of people who have many common tools at their disposal. That set of people will not have all the same tools at their disposal. They may share few if any common values. They may and often do live very different kinds of lives and have very different sorts of priorities. By identifying a "culture" we are at most asserting that everyone in this group of people will use some culturally constructed tools that most of the other members of the group also use. They need not all have any common element at all.

Why then, do people place such heavy and personal importance on culture, if it is really a very fuzzy-edged, flexible, changeable thing?

I think it's because culture determines so much of our reality. Although language doesn't restrain thought or action, culture can. You can't build a good, solid house without the necessary tools. Although we could imagine a culture that doesn't need hammers or lumber to build a solid house - the science fiction readers among you can imagine using nanotechnology to just grow a house without any carpentry or masonry at all - we can't imagine building one without any tools at all, or at least we couldn't build one that would be very solid. Without the culturally constructed tools of home-building, the kind of solid, reliable housing that forms a part of our real lives just isn't real.

This same sort of logic applies to more symbolic tools as well. Consider a cultural artefact like property. Property forms a very important part of our reality. The cultural limitations we place on what is mine and what is somebody else's touch almost every element of our lives. It is such a potent symbol that we actually send people to prison for decades for refusing to acknowledge it as a real thing.

To completely lose your culture is to be plunged into an incomprehensible and unreal world. The things that enable you to develop freely disappear. Furthermore, much of what you are is only accessible through the tools that your culture makes available to you. For your culture to die means that a part of you - a large part of you - is also dead.

That is the main injustice done in the death of a culture. It deprives people of the tools of their free self-development. This injustice is particularly acute when it is done quickly, and when it happens quickly, it is always done incompletely.

Most of the world's minority aboriginal communities offer good examples of exactly this problem. People in these communities often have access to a set of cultural tools which are no longer well adapted to the world they live in and only limited access to better suited tools. Then, the ill-adapted tools can disappear without ever being replaced by better ones. The cultural tools that they have taken as their own are missing in the dominant community, but they don't serve them very well within their own community.

Language plays an important role in all this. Language is a sort of meta-tool. Not only does it have instrumental value as a tool of cognition, its structure and categories also reflect the other tools accessible to its speakers. Culture is reflected and supported in the way people communicate. When the cultural artefacts that support life and development change, the language must change with it, just as Inuit developed vocabulary and structures for discussing snowmobiles when they became a part of their lives.

The problem for language activists arises when the cultural artefacts change, but the language doesn't reflect the changes.

In a strictly monolingual community, this can't happen. New artefacts may come into the lives of the language's speakers, but they will necessarily create ways to express them. They may well borrow those words from another language - sometimes borrowing vast numbers of words and structures - but once they have done so, those words are their own. The language will change with the culture, but it will persist. Its traditions and texts, and its way of categorising the world, will not be totally lost. Instead, it will fade over time as the lives of the languages speakers are transformed.

In a largely bilingual community, the risk that a language will no longer reflect the changes in the lives of speakers is much larger. My mother's native language, Mennonite Plautdietsch, is dying in part because so much of day to day life can only be discussed in that language by relying on the English knowledge of the speakers. There is no vocabulary in Plautdietsch for describing computers, traffic jams, government bureaucracy or even discussing the news beyond the farm report. Were there still any large number of monolingual speakers in Canada, who had adopted those words themselves and could keep up their end of the conversation in Plautdietsch, this would not be a problem. Plautdietsch would still be a productive language, adapting to the transformation of its speakers lives. Instead, the reality of the few remaining Plautdietsch speakers' lives largely no longer takes place in Plautdietsch, even when they still speak the language.

It is this fear that motivates a very diverse array of policies both for cultural and linguistic preservation and should inform language preservation efforts.

In my second post, I outlined an economic argument why large, well-established, sustainable language communities ought not to be suppressed. This argument translates well into a self-development driven framework. It is really not much more than a kind of enlightened self-interest translated into the language of free development instead of economics. My capacity for self-development is highly contingent on the world around me, and particularly on the ability of others to develop themselves. This is one of the key lessons of the notion of distributed identity I developed in the last post.

Simply recognising that language has instrumental value to people - that it is a very important tool of self-development - suffices, in my opinion, to justify linguistic tolerance for any community large enough and established enough to sustain itself and offer a reasonably complete set of options to people. This means more than not punishing people for speaking their language, it means giving them the opportunity to construct their own institutions and to remain attached to them.

Let me offer a more concrete example of how this works: Consider the effect of scientific work on individual self-development. It should be fairly self-evident that scientific work can, and generally does, empowers people quite remote from the work itself. I suppose there might be some real Luddites out there who disagree, and one might find specific instances where scientific work had the opposite effect, but otherwise I would think that most people would agree that science is, on the whole, a good thing even for those with little interest or awareness of science.

Now, imagine a child growing up in Tijuana who has a real gift for math and science - an Einstein or a Newton who can radically change the world for the better. Unfortunately, he has very little gift for languages. But, he grows up in Tijuana speaking Spanish and attending Mexican schools. There, his teachers recognise his abilities and encourage him. He never does well in English classes, but his overall intellectual abilities are enough to get him a scholarship to a major Mexican university and eventually to a PhD at an elite American university. He develops some knowledge of English by then, but he has a hard time communicating clearly. Still, he's obviously brilliant, his work is well-received, and if his English isn't so good... Well, so his papers get ghost-written. It wouldn't be the first time in the history of science. He is still a productive person, and in the end he does work that changes the world for the better.

Now imagine this same child growing up over the border in the barrios of San Diego. He masters a bit more English when he's young than he would have in Mexico, but he still has no gift for languages. Like many Latin American children in the US, his English is never fully fluent, and his limited ability to fully express himself restricts his success in school and in other areas of life. As a result, he has difficulty in all his classes, not just English, and his teachers assume that he just has no intellectual gifts at all. He drops out of high school instead of flunking and ends up flipping burgers at McDonald's.

We can easily see how the main character of this little dual narrative has lost out by being in an English-only society, but what I think people really need to understand is that everyone loses out in this case. My capacity for self-development is contingent on his. His ability to access the tools he needs enhances my access to the cultural tools I need, even when we don't need the same tools. When he loses, I lose.

This bears a more than coincidental resemblance to "the free development of each is the condition for the free development of all." Unlike the original author of those words, I assert that this is not something the world must become, but that it is how it already is and always has been. This same little narrative applies equally to the man who might be a plumber but is instead a day labourer, or to the man who could run a shop but instead sweeps the floors. Whenever a person might have more freedom of development than they have, I have less freedom of self-development than I could have. If someone can be more productive and isn't, I lose from his lack of productivity. Whenever someone has less education than they would like, I suffer from ignorance, because there is a cognitive resource that might be available to me than I can not use.

Although this kind of argument is not novel, I have never seen anyone apply it to multiculturalism or multilingualism. It is, to me, a sounder argument than assigning an intrinsic value to diversity.

This same argument limits language rights claims where we can't so easily decide whether a community is large enough to be sustainable on its own. Instead of first looking to see if there is a historical injustice to remedy, we should ask if there are individuals who are presently suffering from a lack of freedom of self-development that we can link, at least in part, to language endangerment.

I think we can identify such a community in many other cases, particularly aboriginal communities. Although the problems most such communities face can not all be traced to language per se, we can trace a great many of them to the lack of a society which which offers them a full set of cultural tools for their self-development. Helping them to construct such a society, at least where such efforts are possible, is a better solution than integration, and their language can be an important tool in attaining that goal.

Canada has made some moves in this direction, as has New Zealand. The US and Australia are also moving somewhat tentatively towards reconstructing aboriginal societies instead of imagining that integration will solve the problem. The record of these kinds of efforts is mixed, and I think a Vygotskyan treatment can help explain some of the reasons why.

I think the grand mistake of many cultural and linguistic policies is that they focus on preserving the past culture of a community instead of trying to build a future one. I don't want to say that the past is without value and should be unceremoniously abandoned, but I see a lot of community cultural projects that revolve around things like collecting oral narratives or teaching children the history and some elements of the traditional religion of their own communities. These things are valuable, but they are not programmes which serve to keep a culture or a language alive.

Instead, I think these sorts of efforts should focus on avoiding Péire Pessamessa's complaint about Occitan. We cannot prevent cultures from changing and adapting to the circumstances of life and we should not want to. Sometimes, cultural and linguistic preservation projects have to start with people reclaiming their past, but it is a doomed effort if they can't adapt what they find to the present. It is crucially important to develop cultural tools which people can identify as their own and which serve them now.

This is why a language preservation project should place a great deal of emphasis on ensuring that the elements of daily life are available in the language we seek to preserve. There needs to be a basic vocabulary for computers and for government services in that language, and it needs to be a consistent vocabulary that speakers know and understand. It must be possible to communicate without relying on speakers' knowledge of another language.

Inuktitut is one case where these kinds of efforts are presently taking place. With the establishment of Nunavut as a government and the creation of fairly important community-controlled institutions, more and more kinds of materials are available in Inuktitut. This web page, for instance, has a glossary of statistical and demographic terms in Inuktitut for use in government publications. Some are borrowed words, many are created terms, but they are all Inuktitut. By doing this - working to provide a consistent, usable, known vocabulary for concepts which were not present in traditional Inuit culture - they are ensuring that reality continues to happen in Inuktitut. This sort of planned linguistic innovation is utterly essential to the survival of endangered languages.

It is this same thinking which motivates French language policy. French is not endangered in the way that Inuktitut is. There is no need to rely on people's knowledge of a foreign language to conduct your day-to-day life in French. However, it is growing harder to conduct scientific research or high-tech engineering in French without relying on people's knowledge of English.

French language policy is motivated by this fear that many essential cultural elements can no longer take place in French. It is not an effort to shelter French society or the French language from change. In fact, it is most often exactly the opposite. French language policy institutions actively create and promulgate new terms and new expressions in an unending effort to keep up with cultural change. The motivation for these kinds of policies is the desire to ensure that the cluster of tools supported by the French language can continue to serve people, so that French-speaking people can develop themselves as freely as possible.

That is what has been denied to languages like Occitan and other endangered languages. The irony that France on the one hand defends the national language while having a long history of repressing regional languages is not lost on me.

I have already proposed that the freedom of self-development be the metre with which we judge our institutions, and I believe that it should be the uniform goal of all language policy. Language policy should enhance development rather than restrict it.

That means that we must not gain linguistic security by denying people access to other languages that can still further enhance their freedom of self-development. It will probably never be possible to conduct high-energy physics in Inuktitut. If there is some young Einstein growing up in Iqaluit right now, he will need English. Inuktitut-speakers are too small a group to make it possible to do everything without access to a second language, and to make that a goal is unreasonable. However, the ability to conduct as much of one's life as is feasible in the language of one's choice is not an unreasonable goal and where it is present, even universal bilingualism does not necessarily threaten the language's future.

What I reject is the argument that the liability for a historical injustice done to the speakers of a language is owed in any sense to that language. It is only owed to people, and preserving a language is only merited when it serves to restore a capacity for self-development unjustly denied to real, living people. This replaces the idea that some time limit or immigration status should separate justified language rights claims from unjustified ones. When we can't identify a real, existing community that is suffering because a language is threatened, we have less justification in taking radical measures to save it, regardless of the injustice done.

This does not eliminate the hard cases. There are communities where cultural identity persists, but the language is already extinct. Israel has managed to bring a language back from the dead and turn it into a vehicle well suited to communication in the modern world, but Israel had resources that are not available to most communities. Besides, in terms of language preservation, Hebrew has done as much harm as good. Hebrew thrives, but Ladino will not survive another generation and Yiddish - a language with a rich history and literature which only three generations ago had millions of speakers - will probably not reach the end of this century.

Instead, activists on behalf of those languages might entertain a number of options to empower their communities. One option is to opt for a related but less endangered languages. I have hopes that the various Cree/Montaignais/Innu speakers, perhaps in conjunction with smaller communities whose languages are more endangered or extinct like the Abenaki, the Micmac and the Ojibwa, might adopt a single common language as a superior vehicle for their cultural self-preservation than adopting English or French. Another is to give up on language and attempt to develop unique cultural tools using English as a medium.

As appealing as the personality principle is, realistically there will always need to be a territorial element in language policy. Inuktitut is gaining ground because Nunavut is an institution with the power to enforce language policy on a fixed territory. French has in large part recovered in Quebec, and may be gaining ground in New Brunswick. Except in the Ottawa Valley - which is both near Quebec and the seat of the bilingual federal government - it is clearly endangered in the rest of Canada.

Territorial solutions require a fairly high level of political will. This is the problem that has plagued Ireland and the UK. Only 18% of Welsh residents genuinely speak Welsh, and somewhere between a seventh and a quarter of Irish residents speak Irish, depending on the standard of Irish you apply. One can justify measures to restore those languages on the grounds that they are necessary to ensure the freedom of self-development of those communities, but as time goes by it gets harder and harder to make that case. Precisely who is discriminating against the Welsh or Scottish Gaelic speakers, or for that matter the Irish in the Republic?

In Ireland and Scotland, much of the blame for this decline rests with the very state that was supposed to be preserving their ethnic and cultural identity. Scotland has been effectively bilingual for a very long time, and it was the Scots English speakers - particularly the Scottish lords - who were most instrumental in pushing Scots Gaelic out of the highlands and into the outer islands. In Ireland, at the time of independence there were some half-million regular Irish speakers. Now, the figure may be as low as a 50,000. The Irish Free State did not press for an aggressive language policy at a time when it was more likely to have been publicly accepted, and now suffers the consequences.

Ireland and the UK have begun to move towards a more narrowly tailored territorial solution. In Ireland, language law recognises one area - the Gaeltacht - as more Irish than others. It is the part of western Ireland where the majority of the population still speaks and regularly uses Irish. Unfortunately, this area includes a great many English speakers, and the Irish state has not been willing to make Irish mandatory for business and government in that area in the way that, for example, Quebec and Catalonia have. In the UK, the Outer Islands in Scotland have had official bilingualism for a little over 30 years. They still do not have fully Scots Gaelic schools. Bilingualism is expanding to areas like the Isle of Skye as well. Wales is now actively promoting the Welsh language, and the percentage of the population that speaks Welsh is now roughly stable. However, like in Scotland and Ireland, the area where the language is genuinely used in day to day life is relatively isolated.

I think the kinds of language policies that can serve these cases have to be somewhat narrowly tailored. I think all three cases could be well served by identifying those areas where reasonable majorities of the population still speak Celtic languages and enact the kind of language charter that has served well in Quebec. This should happen in conjunction with a tax break for business investment in those areas and the preservation of weaker language protections elsewhere. In all three places - Scotland, Ireland and Wales - there is at least significant political support for language promotion. By identifying the areas where the language is still strong and using moderately coercive measures to keep them that way, it is still possible to keep them from extinction.

Many English monolingual Canadians now identify French as part of their own national identity, so much so that they will gladly pay for quite costly language support programmes in areas with very small French-speaking minorities. Anglophones who do speak French often take pride in it, as a way of showing how loyally Canadian they are. I see no reason why the same thing could not happen in Wales, Ireland or Scotland in conjunction with a narrow and territorial language policy affecting those areas where the language is still fairly strong. A language can constitute an element of identity even to people who don't speak it.

In short, I am recognising that there are times and places where a genuinely coercive official languages policy, instead of a merely pragmatic one, is at least reasonable. There are several factors that I think have to come into play in this sort of instance. First, it must enhance the freedom of self-development for some real group of people without unduly restraining it for others. The language that this policy is trying to save must have the political support of some population willing to pay a price for it. If the Irish want to preserve Irish and are willing to subsidise the Gaeltacht in order to make that happen, that is acceptable to me. Second, the policy should not undermine a language that isn't secure in some other area of the world. An English speaker living in an all-Welsh village still has access to hundreds of millions of English speakers elsewhere in the world. An Abenaki speaker in Quebec should not be forced to use French over Abenaki. This sort of policy must also not restrict the freedom of movement either of those inside or outside of the designated territory and it should be able to offer the most basic kinds of services to speakers of the areas most important minority languages. People who wish to live in a language protected area should recognise that they ought to learn their neighbours language not out of altruism but for their own sake, and the people who already live in them should recognise that isolating newcomers from basic services until they learn the language harms them as much as it harms the immigrants.

It is especially important that racial and ethnic criteria not play a part in language policy. New Zealand is now trying to encourage non-Maori to learn the Maori language precisely because their language and culture will survive better by integrating willing outsiders than by demanding a set of rights predicated on ancestry and ethnicity. If Ireland's Gaeltacht is to be Irish language-only country, that should not prevent a Gujarati immigrant from moving in, learning Irish, and being an equal member of the community. Mixing language, race, ethnicity, nationality and religion is almost always a bad thing.

That is much of what makes a defence of language rights different from "separate but equal." First, it isn't really equal. By identifying that different circumstances call for different measures, a good set of language policies is never truly equal. It shouldn't be equal, it should be optimal. It should offer the most freedom to the most people.

Second, it isn't separate. I expect and I want people to move around freely across language lines, and to learn the languages of the people they are most likely to need to communicate with. I don't want one group of people living in their community and another group in a different one. Good fences may make good neighbours, but actually interacting with other people is a lot more fun and a lot more stimulating.

I want to end this post with that vision of a multilingual world, where people have both language security and the resources to keep language from becoming a barrier to development. That is what language policy should be for.




1 Quote translated by Scott Reid, Alliance MP for Lanark-Carleton. I came across it in his book Lament for a Notion. I do not endorse Scott Reid's politics, but I don't have a copy of Leclerc's book on hand.
 

Wednesday, August 27, 2003
 
Yet another reason to move to MT

For whatever reason, comments are still dysfunctional today. You can see them by turning off Javascript. If the problem hasn't gone away by evening, I'll... I dunno... do something, I guess.

Update: I've identified the problem - it's Enetation's fault. They've got a syntax error in their Javascript.

Further Update: Trying to fix the problem. I'm thinking maybe it's time I invested in a domain and just went over to MT.

Temporary Fix: You can't see how many comments a post has, but at least you can get to the comments.
 

Tuesday, August 26, 2003
 
Short delays

Well, it looks like I am once again behind schedule. There is still a fourth post on language rights coming but it's still in bit and pieces. With some hope, it will go up Wednesday evening my time - Wednesday during the day for my North American readers and Thursday morning for my Pacific Rim readers.

The problem is that my boss is suffering under the delusion that just because he pays me, I'm expected to do what he asks. :^)

In the meantime, I strongly recommend reading the comments to the language rights posts so far (if you happen to be lucky enough to actually see them when you load the page) and the posts of the folks who have linked in. Besides Jacob Levy, Matthew Yglesias and Language Hat, I see that Ampersand at Alas, a Blog, Prometheus 6 and Debitage have all linked in. All have some good points that I need to get to.

Also, I have recently had some research on ICAO English and air traffic control come across my desk. There's some interesting stuff going on there. The air travel industry is more or less the only place where global, mandatory functional monolingualism has genuinely been tried, and the results are awfully mixed. I think the linguistic troubles of the air industry can inform language policy issues in general, and I have in my head the outline of a post on it, which I will get to as soon as I can.
 

Sunday, August 24, 2003
 
Proof positive that I'm a hypocritical bastard

This is an extract from what I've been working on all weekend:

Quality control and measurement has long been considered a hard problem in translation and there is no well-established body of theory or practice treating this problem. The principle barrier to addressing quality issues in translation is the difficulty quantifying quality. There have been efforts in the past to use human readers and revisers to score translations, but the subjectivity and inconsistency of these measures, as well as the added expense of a human reader, has rendered these methods impractical on any large scale. This project takes on the challenge of developing effective and consistent quailty metrics for translation that can be cheaply deployed, and building processes to enhance quality and productivity using those metrics.

At present, the state-of-the-art in translator productivity software includes a toolkit of terminology management, translation memory and machine translation software. These tools are most frequently owned and controlled by the translators themselves, who must spend time maintaining them. Translators judge for themselves how much time and effort to put into their tools in order to optimise their own productivity, and because translators are most frequently paid by the word rather than the hour, they alone benefit from whatever productivity gains their tools offer them. In recent years, many translation firms have moved to centralise their productivity tools in order to further enhance their value and translation tool vendors have started offering more centralised software suites.

However, in order to realise gains from these efforts, translation firms have had to modify their payment schedules. Translators see their pay per word reduced on the basis of an estimate of how useful these centralised resources are in translating an element of text. Generally, a firm may adjust translator pay in one or more of a small number of ways. They may ask translators to pay for the terminology they receive from the translation agency. They may lower translator pay when they receive perfect or high quality matches from a translation memory. And, they may treat machine translations in the same way as translation memory matches, usually offering a single reduced rate for some or all of the text segments that have been automatically translated.

We are unaware of any research justifying adjusted translator pay scales on the basis of actually measured productivity enhancement. These schemes tend to be a matter of hit and miss, based on anecdotal evidence and negotiations between translators and translation agencies. Furthermore, in the past when translators managed their own productivity tools, they could presumably best judge how much gain they were actually getting out of each tool and divide the time they spent maintaining them appropriately. By centralising these productivity tools, this feedback loop has been broken. It is impossible to estimate how much productivity translators gain from particular tools, nor how much labour savings is derived from a given amount of maintenance effort. Without this knowledge, it is possible to actually reduce productivity by misdirecting labour. There is anecdotal evidence that companies may well be spending the most labour maintaining the tools that least increase translator productivity. The hit-and-miss nature of these tools almost ensures inefficient central management.

The knowledge that we hope to acquire through this project is a way to restore the feedback from the translator to the productivity tools. We are not aware of any other effort oriented specifically towards this goal. By measuring translation quality consistently and cheaply, we are able to determine how much impact our central resource management policies have on translator's work, and by more accurately estimating the value of our central resources to translators, we can adjust their pay scales in a manner that genuinely reflects their productivity gains.

That's right. I'm working specifically on ways to lower someone's pay. I know there are a number of translators reading my blog who will be appalled, but not any more than I am. At least, if I get my way, we'll only be paying you less when we actually do something useful for you.