Махновщина

A rambling discussion of language rights in education

2010-06-01T18:06:00.004+01:00

Education is one of the primary arenas for the enactment of language rights. Negative education policies can go to the core of those problematic issues which the language rights movement seeks to redress: the oppression of and discrimination against minority groups and immigrants; the degradation and erosion of traditional, disempowered cultures; the maintenance and entrenchment of existing social inequalities through the medium of language.

Many language rights have relevance in education. One of the simplest (although not the earliest to be expressed), the right to speak one’s own language, has historically been directly broken most frequently in the context of schools, as oppressive regimes in countries ranging from the United States to Norway sought openly to eradicate the cultures of their minorities by preventing the transmission of these cultures to children. Related rights, such as the right to take part in cultural life and the right to associate with members of one’s cultural and linguistic communities of origin, have also been violated through these and similar educational policies. Nevertheless, the right to speak one’s own language in school stands as a singularly symbolic and central right in the protection of threatened cultures and persons belonging to minorities: it is by the violation of this right that cultures and peoples may be most efficiently and thoroughly discredited, obscured and forgotten.

For this reason, other rights have been formulated with special relevance to education. These stem from the right to education itself but go much further, including rights to an education which is directed at the development of the whole personality, to an education in one’s culture and language of origin, and to an education which develops respect for one’s parents and culture of origin.

Other educational rights are of indirect relevance to language as they must be mediated through language. The right to an effective education – or indeed an education at all – is reliant on an educational system that attends to the needs of the child. In the case of children who are not speakers of the majority or standard language, such needs clearly include special attention to their education in that language, so as to allow them the social mobility and opportunity afforded by the education system to other children. However, this point goes further: research over the last fifty years has conclusively shown that education of such children must be bilingual if it is to be effective, teaching not only in and about the majority language but also in and about the minority languages spoken natively by pupils. Where monolingual, non-native-language education has been shown to be psychologically, educationally and socially detrimental to the individual, it has been shown that bilingual education can be highly beneficial.

Various different approaches have been used in the discussion of language rights. The earliest, Linguistic Human Rights, views language rights as a subset of human rights and having all the same properties as these rights: universality, inalienability, an individual focus, and a focus on the nation-state as duty-holder. However this approach has been criticised on various fronts. In the arena of education, it can be observed that the human rights approach to language rights – at least traditionally – has offered too simplistic a model of language in society to be fully effective in all cases. The liberal approach has typically assumed a model of language and society that descends from the conception of the modernist nation state (from which human rights discourse itself descends): languages are discrete, abstract entities; they are strengthened and improved by being written; states should coincide, where possible, with language communities, and should be symbolised by their national languages as a central pillar of national unity. Although, of course, the a language rights discourse will never advocate the progression of monolingualism, nevertheless this ideology may represent the background to its approach. But it overlooks many issues: standardising a language and devising a written form for it may reproduce the situation found before language rights were considered, creating new minorities and new non-standard forms to be stigmatised; there may be no clear boundary, either linguistically and/or in speakers’ conceptions, between the stigmatised language variety and other varieties – indeed, the minority language may not be considered a language at all – and so it may be very difficult to persuade speakers of the wisdom of treating it like a majority language; standardising and writing a language puts it into the same arena as dominant languages, yet in this arena, as economic and cultural capital, it may be unable to compete; speakers’ identities may be based on elements of their language ideologies that simply do not match the scientific approach taken by language rights activists, and so be damaged by language rights activism.

Other approaches exist also: the ecological model of language diversity proposes the metaphor that languages are species, language shift is extinction or death and dwindling languages are endangered. This may be a powerful tool in persuading others of the value of language rights activism, but has little specific relevance to education, and even in its more general application may also have faults: the metaphor taken too far may lend opponents arguments, such as proposing that languages undergoing shift are those ‘unable to compete’ and so in some sense destined to die.

Education is of very wide relevance across all societal groups. Nevertheless, language education and language rights in education are of course of relevance to certain groups more than others. Firstly, it is generally of particular relevance to minorities. No legally accepted definition of minorities exists, although one (non-legally-binding) possibility was proposed in Capotorti 1991 for the UN:

A group “numerically inferior to the rest of the population of a State, in a non-dominant position, whose members – being nationals of the States – possess ethnic, religious or linguistic characteristics differing from the rest of the population and show if only implicitly a sense of solidarity”

Where such groups possess linguistic characteristics differing from the rest of the population (which, even where it is not their defining feature, is likely to be the case) language rights in education are of particular relevance to them. As has been mentioned, children are not well-served by education in a language other than their native language, and education not tailored to their needs is likely to be to their own detriment as well as to the detriment of their culture. Such inappropriate education in the long term produces high school drop-out rates, low rates of academic achievement, high rates of unemployment and high rates of reliance on state benefits: all in all, high rates of rejection of productive interaction with and operation within the state (this has historically been demonstrated, for example, by the treatment of Native American groups and African American groups in the United States).

Also of relevance are particular types of minorities. The Native American groups mentioned about constitute an indigenous minority: a distinct ethnic-cultural group with a history predating European settlement of their country. Indigenous peoples, although they constitute a tiny proportion of the world’s total population, represent some of the world’s most oppressed and discriminated groups. Also of relevance are national minorities. These are defined by the 1998 Oslo Recommendations as:

“those situations involving persons belonging to national/ethnic groups who constitute the numerical majority in one state, but the numerical minority in another (usually neighbouring) state”

National minorities differ from migrants as they may be citizens of the country in which they live (and in which they are a minority); they may receive some more protection of their rights as they are associated with another powerful duty-holder, but this may equally work to their detriment. Where other minority groups are usually – by their very nature – not threatening to the hegemony of the majority, national minorities may represent a group which – on a larger scale still – is the majority. Such cases are found, for example, in the Russian minorities in various states which were once Soviet Satellite states: although ethnic and linguistic Russians are a minority in the context of these states, they may come to represent Russia, a powerful and potentially dangerous neighbour, and so be subject to discriminatory attitudes.

Although no legal definition of ‘minority’ is available, ‘minority language’ has been legally defined: the European Charter for Regional or Minority Languages (1992) states:

“’regional or minority languages’ means those languages that are: i. traditionally used within a given territory of a state by nationals of that state who form a group numerically smaller than the rest of the state’s population; and ii. different from the official language(s) of the state; iii. it does not include either dialects of official language(s) of the state or the languages of migrants”

This definition covers various many different cases of groups who would otherwise be at severe risk of discrimination, and the rights laid out in this and other treaties have been instrumental in enacting significant change for these groups.

It may be noted that the definitions given above have all excluded migrants: these individuals are at particular risk in the arena of language rights and education. They are usually representative of much smaller local communities and so have smaller support networks other than the state itself. They, like the other minorities considered, often speak different language varieties and so are in need of special educational measures if they are to fully benefit from the education offered to their peers. However, due to their small numbers the cost of compliance with their language rights and educational needs would be extremely high per person. Additionally, their status as (typically voluntary) migrants labels them as in some sense less deserving in the view of some political approaches and arguments about indigenous origins cannot be made to support their claims; they are thus more easily swept aside.

Many different models may be found in the education systems of states around the world for good and bad language policy and support of language rights. Educational systems may be roughly grouped into three types: monolingual education, which teaches in only one language; subtractive bilingual education, which teaches in more than one language but with the sole aim of proficiency in the dominant language; and additive bilingual education, which teaches in more than one language because it aims at proficiency in more than one language – at bi- or multilingualism.

An example of monolingual education is found in the US programme of “sheltered immersion”. This is a programme in which non-English-speakers are immersed in English – taught only in English, alongside their English-native peers – but are given additional lessons in English as an L2 alongside. Another quite different example of a monolingual education system was found under Apartheid South Africa: this system taught minorities only in minority languages, effectively restricting access to the dominant language. Both of these examples – and indeed monolingual education generally – are ineffective at producing well-educated, well-rounded students, especially with reference to maintaining social mobility alongside native cultures. The programme of the type described in the US tends to produce individuals who get on in the educational system less well than their peers whilst at the same time lowering their esteem for their own cultural heritage and failing to develop their proficiency in their first language. The segregationist programme of the type described in Apartheid South Africa, by limiting access to the dominant language, highly discourages social mobility.

An example of subtractive bilingual education is also commonly found in the US: in this system, students are first taught in their L1 alongside lessons in English; they are then slowly streamed into English-only education. This type of education, although more effective than monolingual immersion, still suffers from the same criticisms.

An example of additive bilingual education is found in Canada under the title of “Heritage Language Bilingual Education”. This programme teaches minorities in both minority and dominant languages side by side, and is much more successful at producing high levels of bilingualism and highly educated individuals. A very different example was found in the Spanish Basque Country under Franco. This type of educational system, described as a separationist system, was organised by the minority group themselves and was more political in its intentions than educationalist.

Approaches in language rights and their failings

2010-05-30T17:29:00.002+01:00

Over the latter half of the twentieth century and the beginning of the twenty-first, a discourse has developed around the issues of ‘language rights’ and ‘language survival’. These two topics stem from the recognition of two different social problems: in the case of language rights, the oppression and denial of the rights of those who speak minority languages; in the case of language survival, the incredibly rapid levelling and loss of language varieties the world over. Nevertheless these two topics are clearly closely causally related and should share common solutions, and this has united discourse around them. Traditionally most of this discourse has been framed in terms of of the Linguistic Human Rights paradigm, but more recently this approach has been subject to some criticism and other approaches have developed also.

Linguistic Human Rights
The Linguistic Human Rights approach views language rights as a ‘chapter’ in the larger, universalising discourse of Human Rights. It sees discrimination on the basis of language in the same category as discrimination on the basis of sex, gender, race, etc., and assumes that prototypical language rights share the qualities of prototypical human rights. This is by far the most developed and widespread paradigm, but it has also been subject to the most criticism.

One issue with LHR derives from the fact that Human Rights discourse, despite its claims of universality, is historically essentially European. As such it is problematically bound up with other modernist European ideas around the nation-state. Such ideas include linguistically homogenous nations: this highlights the risk that LHR may reproduce the modernist nation-building project in the way that it attempts to institutionalise threatened languages, creating in doing so new minorities and new oppressive language ideologies. Equally, legal rights discourse, both national and international (as international law relies on agreement between nation-states), is framed as a discourse between the individual and the nation-state. Yet the nation-state is ideologically reliant on the ideology of monolingualism, and so is highly incentivised in practice to reject the LHR discourse.

Another related point is that the universality which is central to HR discourse (and so LHR discourse) presents practical problems in translating normative rights into legal rights: rights which are universal are much more costly for governments to uphold than targeted, group-specific rights, and so are less likely to be put into law.

A different criticism of LHR is that it tends to view language as central to ethnic identity: this point is itself central to arguments, representing one of the major motivations to save dying languages and support specifically the language rights of linguistic minorities. It is also to be expected from a discourse which developed out of a discipline of linguistics. However evidence in practice as well as more up-to-date sociological theory belies this: ethnic identities often survive the death or levelling of their associated language variety without great loss of vitality, and it is more accurate to think of ethnic identities as multiple, actively constructed, situated in practice, and not necessarily associated with language. If this is not to undermine the case for language rights, then the centrality of language to ethnic identity must not be ncessary.

A similar criticism arising as a result of the origins of LHR is related to its formal linguistic element. LHR often concentrates strongly on a formal linguistic features which have little to do with how speakers themselves view language. It thus offers no insight into how to approach a situation in which a language variety has died as far as formal linguistics is concerned but still exists in the views of speakers; similarly it offers little insight into how to approach situations in which linguistic definitions of language varieties do not match up to speaker definitions, and in imposing its own definitions may damage speaker identities. At other times LHR may not take into account linguistic facts but nevertheless for political reasons identify linguistically homogeneous and distinct ‘nations’ where none exist, either in practice or in speaker ideologies, which go on to superimpose over and replace existing speaker identities.

Finally, it is unclear in the LHR paradigm how to approach the issue of situations where preservation of a minority language variety may be politically or economically damaging to speakers.

Language Ecology
A language ecology approach, using the metaphor that languages are species and language diversity is biodiversity, refers to language shift as language death or extinction, and dwindling languages as ‘endangered’. It builds on the existing widespread nature and considerable rhetorical power of biodiversity discourse. This is also a widespread and well-developed approach, and in practice lends its terminology to much of the language rights discourse. However it, too, presents some problems.

One issue is that the metaphor itself is no more than a metaphor: languages are biological only in the sense that the language faculty is biological; they are not genetically inherited. It may be tempting to use biodiversity arguments in favour of language preservation, but language death does not actually reduce biodiversity, and arguments for the preservation of biodiversity cannot automatically be used in favour of the preservation of linguistic diversity. When taken too far, this metaphor can generate very weak or even fallacious arguments in favour of language preservation.

Furthermore, the metaphor can be turned against the language survival discourse: if languages are ‘species’ and species die naturally, then it is natural that languages die too. In addition, species go extinct as a result of failure to compete, and so are in some sense fated or deserve extinction: thus the same can be argued about dying languages. This is, of course, again taking the metaphor too fast: it is not languages but speakers and cultures that compete.

Another problem with this approach, in common with the Linguistic Human Rights approach, is that it conceptualises languages as objects in themselves. This obviously gives an inaccurate picture of reality and can result in overlooking the real socio-political complexity involved in speaker communities and identities.

The metaphor also tends towards the viewpoint that all change is bad and, as far as possible, species-in-habitats should be returned to their historic states. Again this is an impractical and overly simplistic approach, and fails to recognise the fact that linguistic practices are cultural practices and express cultural meanings which do not die when languages die, although they may change. However this points towards a potentially more positive approach: considering language ‘habitats’ – the sociocultural spaces in which linguistic practices exist and are valued – and how they might best be preserved even whilst elements of material culture necessarily change.

The postmodernist approach
Most recently the postmodernist (or emic) approach has developed. This approach focuses more on speaker conceptions of language and language identities, on performance and context, and on a wider and more pluralistic conception of ‘speech communities’. It too is not entirely without problems, although as it develops out of criticism of earlier approaches it is commonly accompanied by the admission that elements of an etic approach are likely necessarily alongside the emic.

The primary issue with this approach is that it may highlight plural identities and blur linguistic boundaries so much so that it leaves little concrete around which champions of language or minority rights can rally: the overly simplistic conceptualisations of earlier approaches are far more viable tools in the political arena than the very complex picture that an emic approach can paint.

One secondary issue that can be highlighted is that linguistic identities and ideologies, on which the emic approach builds its conceptions, are not invariable, and change anyway through contact and over time: the extreme cultural changes of the modern era merely accelerate this process. Thus it might be argued that it is not clear how such an approach offers a reason to strive for the preservation of languages at all.

A brief introduction to Optimality Theory in phonology: Part 2

2010-05-23T12:40:00.008+01:00

Prosody in OT and Generalised Alignment
Here I will briefly introduce some constraints relevant to the analysis of prosody in OT – as in much of the rest of this introduction, I will assume a prior knowledge of the linguistic issues involved.

The first key constraint in analysing prosodic structure is PARSE-σ: this constraint states that all syllables must be parsed into feet. A language in which PARSE-σ is highly ranked may allow degenerate feet rather than allow some material to remain unparsed, may delete material rather than allow it to remain unparsed, or may epenthesise material so that legal feet may be formed. Note that PARSE was used in the earliest OT literature with a quite different meaning – in the this literature it was a constraint against deletion, equivalent to MAX in modern OT.

Various constraints are relevant in determining the form of prosodic feet. The first is FOOTBINARY (FTBIN): this constraint states that prosodic feet must be binary (disyllabic or bimoraic depending on the language in question). A language in which FTBIN is not highly ranked may allow degenerate feet, or may have only a single foot (and only a single stress) per word. Also relevant are constraints to determine the headedness of feet: these constraints usually take the form FOOTTYPE_TROCHEE and FOOTYPE_IAMB (FTTYPE_TROCHEE and FTTYPE_IAMB), and state that feet must be trochaic/iambic respectively. Some authors instead use simply TROCHEE and IAMB.

Sensitivity to stress is controlled by the constraint WEIGHT-TO-STRESS-PRINCIPAL (abbreviated WTS and WSP by different authors). This constraint that heavy syllables must be stressed.

Extrametricality in OT is dealt with by the constraint NONFINALITY (NONFIN); this constraint states that there is no metrical structure on the right edge of the word (Rocca et al 1999:609).

Now we come to the most important constraints to be introduced in this section: alignment constraints. Generalised Alignment theory was proposed in Prince & McCarthy 1993, and provides a scheme of constraints to control the position of elements with regards to one another. These constraints are not only used to control the position of phonological elements, but occur also in OT analyses of other areas. The constraint, ALIGN, takes the following form:

“Align (Cat1, Edge1, Cat2, Edge2) =_def
Cat1 Cat2 such that Edge1 of Cat1 and Edge2 of Cat2 coincide.
Where Cat1, Cat2 ProsCat GramCat
Edge1, Edge2 {Right, Left}
(Prince et al. 1993:2)

In prose, this means that a given ALIGN constraint states that an edge (the left edge or right edge) of every instance of a given category must align with an edge of some instance of another category. For example, we might see:

ALIGN(Ft, L, PrWd, L)

This constraint states that the L[eft] edge of every F[oo]t must align with the L[eft] edge of a Pr[osodic] W[or]d. As can be seen, this is the equivalent of stating that syllables are parsed into feet from left to right instead of from right to left.

I will now undertake a brief analysis of stress placement in Latin to demonstrate these constraints. Firstly, here is some data to demonstrate Latin stress placement (mostly taken from Rocca et al. 1999:629):

amí:cus
agricolá:rum
vólucres
símula:
magnópere
refé:cit
agrícola
pepérci
magíster
ámo:
ré:

From examining these data, we can make some generalisations about the placement of Latin stress:

• the weight of the final syllable makes no difference;
• if the penultimate syllable is heavy (either because it contains a long vowel or a coda consonant), then it is stressed;
• if the penultimate syllable is light, then the antepenultimate syllable is stressed.

How can this be modelled in OT? It is clear that stress falls within the three syllable window and so the word is right-headed; it is also clear that syllables are parsed into feet from right to left, as there is no difference in the placement of stress between words with even and odd numbers of syllables. Thus we can posit the following constraint:

ALIGN(Ft, R, PrWd, R): The right edges of all feet are aligned with the right edge of some prosodic word.

This also points towards trochaic feet:

FTTYPE_TROCHEE: All feet are trochaic.
FTBIN: Feet a disyllabic.

It can also be seen that the final syllable must be extrametrical, as (1) stress placement is sensitive to weight but a heavy final syllable is not stressed, and (2) when weight is not a factor, it is the antepenult which is stressed. Thus we can add another constraint:

NONFIN: “No head of the prosodic word is final in the prosodic word.” (wording from Prince & Smolensky)

As we can see that stressed monosyllables do occur, DEP must outrank NONFIN so as to avoid the epenthesis of an extra syllable to avoid final syllable stress; it must be outranked by PARSE-Pr[osodic]W[or]d so that monosyllables are parsed:

DEP, PARSE-PrWd >> NONFIN

However we can also see that NONFIN must outrank ALIGN(Ft, R, PrWd, R), as NONFIN prevents final syllables in multisyllabic words from bearing stress:

DEP, PARSE-PrWd >> NONFIN >> ALIGN(Ft, R, PrWd, R)

We now come to weight sensitivity. We must add the WSP constraint:

WSP: Heavy syllables are stressed.

This constraint must be outranked by NONFIN (otherwise final heavy syllables would be stressed), but must outrank either FTTYPE_TROCHEE or FTBIN or both, as penult stress must be parsed either (σσ́)σ (violating FTTYPE TROCHEE) or σ(σ́)σ (violating FTBIN). As FTBIN must be violated in the case of disyllables anyway, it makes sense to suggest that all Latin feet are trochaic and WSP (and FTTYPE TROCHEE) outranks FTBIN. Note that as we known that NONFIN outranks WSP and that DEP and PARSE-PrWd outrank NONFIN, we can conclude that DEP and PARSE-PrWd also outrank WSP: this is known as ranking transitivity.

DEP, PARSE-PrWd >> NONFIN >> {WSP, FTTYPE_TROCHEE >> FTBIN}, ALIGN(Ft, R, PrWd, R)

Now lets see how this produces some of the forms in the data above.

/volucres/	PARSE-PrWd	NONFIN	FTTYPE_TROCHEE	FTBIN	ALIGN(Ft, R, PrWd, R)
☞(vólu)cres					*
(volú)cres			*!		*
(vó)lucres				*!	**
vo(lú)cres				*!	*
vo(lúcres)		*!
vo(lucrés)		!	*
volu(crés)		!		*
volucres	*!

/simula:/	PARSE-PrWd	NONFIN	WSP	FTTYPE_TROCHEE	FTBIN	ALIGN(Ft, R, PrWd, R)
☞(símu)la:			*			*
(simú)la:			*	*!		*
(sí)mula:			*		*!	**
si(mú)la:			*		*!	*
si(múla:)		*!	*
si(mulá:)		!		*
simu(lá:)		!			*
simula:	*!		*

/ami:cus/	PARSE-PrWd	NONFIN	WSP	FTTYPE_TROCHEE	FTBIN	ALIGN(Ft, R, PrWd, R)
(ámi:)cus			*!			*
(amí:)cus				*!		*
(á)mi:cus			*!		*	**
☞a(mí:)cus					*	*
a(mí:cus)		*!
a(mi:cús)		!	*
ami:(cús)		!	*		*
ami:cus	*!		*

/amo:/	PARSE-PrWd	NONFIN	WSP	FTTYPE_TROCHEE	FTBIN	ALIGN(Ft, R, PrWd, R)
(ámo:)		*!	*
(amó:)		!		*
☞(á)mo:			*		*	*
a(mó:)		!			*
amo:	*!		*

/re:/	DEP	PARSE-PrWd	NONFIN	WSP	FTBIN	ALIGN(Ft, R, PrWd, R)
☞(ré:)			**		*
(ré:a)	*!		*
(ré:)a	*!				*	*
re:		*!		*

A brief introduction to Optimality Theory in phonology: Part one

2010-05-20T15:40:00.005+01:00

Optimality Theory was first proposed in the early 1990s, and represented a dramatic departure from previous approaches. It offered a new and quite different conception of language processes and so enabled much more satisfactory explanations of some previously troubling questions. It remains the most vital and widespread theoretical framework in published literature on phonology today, although it has seen many changes since its earliest form.

The central insight of Optimality Theory is a different way of processing changes between underlying and surface forms. Instead of a series of distinct, language-specific rules which are applied without fail in a specific order, OT proposes a series of universal, violable constraints, the differing strengths of which will determine the best possible output. It does away with the rules of previous (rule-based) theories, replacing them with these constraints.

In OT, then, a grammar is conceived of as a series of constraints (which may or may not be universal) which are strictly ordered. The ordering of these constraints provides the basis on which to choose the best possible form of the output, with higher ranked constraints being given a greater weighting. The constraints are thus not absolute: almost all outputs will violate some constraint or other. The form chosen for the output (the best candidate) is the form which violates the least high-ranking constraints.

The process which modifies the input (the underlying form) to generate all of the possible candidates to this process is called the generator, or GEN. The process which then chooses between them on the basis of the constraints is called evaluate, or EVAL.

Markedness and faithfulness constraints
Two distinct classes of constraint are recognised in OT: markedness constraints and faithfulness constraints. Markedness constraints represent linguistic universals: it is generally recognised in linguistic theory that certain forms are more marked (less natural, less common) than others. It is possible to make a series of statements identifying various combinations of features, parsings, or forms as more marked. OT formalises this observation by proposing that there exist universal constraints against these marked forms, which are violated if the marked forms occur in the output. An example of a markedness constraint is NONASAL:

NONASAL: Nothing is nasalised.

This constraint is violated for every segment which is [+nasal].
Faithfulness constraints, contrastively, require that the output be identical to the input (usually – see below). They are violated by differences between the form of a candidate and the form of the input. An example of a faithfulness constraint is IDENT[itity]-I[nput-]O[utput]:

IDENT-IO: All segments in the output are featurally identical to their correspondents in the input.

It can thus be seen that the majority of grammar can be derived by the comparative ranking of faithfulness and markedness constraints: if markedness constraints outrank the relevant faithfulness constraints, then the output will revert to the most unmarked form; if the faithfulness constraints outrank the relevant markedness constraints, the marked elements of the input will be preserved in the output.

To use the above examples, if in a given language NONASAL outranks IDENT-IO then no [+nasal] segments will surface; on the other hand, if IDENT-IO outranks NONASAL, then underlying [+nasal] segments will be preserved in surface forms.

Tableaux
An analysis in OT is thus comprised of a series of constraints and their comparative ranking. To demonstrate that a given constraint ranking will successfully select the correct candidate, it is normal to show the constraints and possible candidates in a tableaux: this is a table with constraints shown along the horizontal and candidates along the vertical; cells are marked to show constraint violations by candidates. From a tableaux it is easy to see which candidates break which constraints and so what candidate a given ranking would select.

Below is an example of a simple constraint ranking and tableaux: specifically, this is a ranking to control vowel and consonant nasality in American English (adapted from Hammond 2003:23). The constraints used are as follows:

NASAL: A vowel before a nasal consonant must be nasalised.
NONASAL: Nothing is nasalised.
IDENT-IO_NASAL(C): All consonant segments in the output are featurally identical to their correspondents in the input for nasality.
MAXIMALITY: All segments in the input have correspondents in the output (MAX).

Their ranking is as follows. Note that the convention is to use “>>” to indicate that one constraint outranks another, and to use a comma to indicate that it is not clear how two constraints are ranked with regards to one another.
MAX, IDENT-IO_NASAL(C) >> NASAL >> NONASAL

The tableaux showing this ranking is below. Note that the optimal candidate – the output candidate – is marked with a pointing hand symbol; that asterisks (*) are used to indicate constraint violations; that an exclamation mark is used to mark the highest ranking constraint violation of each non-optimal candidate.

/ran/	MAX	IDENT-IO_NASAL(C)	NASAL	NONASAL
[ran]			*!	*
☞ [ra~n]				**
[rad]		*!
[ra]	*!
[rn]	*!

“What is the place of language rights in the history of human rights?”

2010-05-19T21:40:00.004+01:00

Human rights as a concept has developed throughout the latter half of the twentieth century; during this time it has expanded to become near-universally known and widely accepted. But what are human rights?

Human rights are conceived of as properties of bodies (rights-holders) which confer duties on other bodies (duty-holders) acting around them. All human rights are conceptually derived from a fundamental, underlying right of a human individual to dignity. Prototypically, human rights are thus absolute, and focused on protecting individuals against states. They are moral as opposed to legal objects.

Language rights cover a various subcategories of human rights: rights conferred or delivered through language, rights that make reference to language, rights to language. Thus the language rights paradigm usually involves choosing to take a view of language as a right; this is in conflict with the common folk-view of language as a social problem, but is consistent with a separate view of language as a resource.

Human rights have developed greatly from their starting point and are often grouped into 'generations' of rights associated with the waves of discussion from which they emerged. The first generation of human rights grew out of the international political flux following the Second World War, but the roots of the concepts underlying them went back further: the constitution of the United States (written in 1787) declares that “[a]ll men are born free and equal, and have certain natural, essential, and unalienable rights”, and itself was influenced by the writings of earlier philosophers such as John Locke and Thomas Hobbs.

As the discourse behind this generation of rights developed out of the chaos and suffering resulting from a breakdown of inter-state relations, the rights themselves were similarly focused on inter-state issues. Rights of this first generation are prototypical human rights: they are (theoretically) universal, inalienable, individual, and focused on freedoms from oppression or maltreatment of the individual by the state. Language is at this point of only limited relevance: it is relevant but unmentioned as the medium for delivery of rights; it is relevant and mentioned in passing as a potential basis for targeted oppression. If it is considered at all, language is at this point conceived of as an inalienable property of an individual.

The second generation of human rights evolved out of the Civil Rights movement of the nineteen-sixties. This generation of rights focused on the protection of minorities and individuals' freedom from discrimination. Language was still not especially central at this point. Again it is relevant to the delivery of certain rights, such as the right to free trial and the right to freedom of expression, and is indeed explicitly mentioned in these contexts. It is also relevant because some minority groups targeted by non-discrimination rights legislation of this period, being associated with no single state, were identifiable primarily by their language; an example of such a group is the Romany.

The third generation of human rights moves in several new directions. It begins to focus more on rights as properties of groups as opposed to properties of individuals (that is, allowing groups or communities to be rights-holders). From this focus develops the distinction of the personality principal and the territoriality principal in the application of human rights: the personality principal applies human rights on an individual basis, whereas the territoriality principal applies human rights to geographical areas according to need.

Rights in this period continued to develop a focus on specific vulnerable, disempowered or minority groups. Important rights include the specific rights of children, such as to education, a development of native culture, identity, language, and respect for parents; these highly specific rights 'trump' more general rights legislation. This generation also sees explicit mention of minority language rights (such as in the 2007 UN Declaration of the Rights of Indigenous Peoples) and (in keeping with the aforementioned development of the territoriality principal) begins to see language as a property of communities instead of individuals. At a late point in this generation discussion also begins of non-state duty-holders (such as international corporations and military insurgent groups).

Thus it can be seen that overall the development of rights over the twentieth century has involved a development from civil rights in national contexts to truly human rights in universal contexts.

Various philosophical issues must be considered in relation to human rights and specifically language rights. Firstly, as mentioned at the outset of this discussion, human rights are typically conceived of as moral, normative and aspirational: they are intended as moral universals to which perfect legal systems should strive, as opposed to being legal objects in themselves. This conception comes with both advantages and disadvantages. On the one hand, as human rights are not legal objects in themselves but represent an aspirational endpoint for the development of legal systems, non-legally-binding rights declarations can still have some effect in this arena where public opinion is sufficiently positive towards human rights generally. On the other hand, the aspirational nature of human rights may offer an argument to duty-holders against passing human rights legislation: if human rights are aspirational then it can be argued that they are not absolute, and that some are less imperative than others; this offers states a justifiable cost of compliance argument against passing rights legislation in given situations.

Another relevant philosophical point is the distinction between positive and negative rights. Negative rights are rights from: they are rights that protect the rights-holder against specific actions committed against them by duty holders. Thus to conform to negative rights, duty-holders need only refrain from performing certain actions. Negative rights date back much further in the development of the human rights discourse and early rights legislation is focused on negative rights. Contrastively, positive rights are rights to: they are rights that grant rights-holders specific goods or services. To conform to positive rights, it is beholden on duty-holders to actively perform actions. Positive rights are less widely accepted in human rights discourse and fewer pieces of positive human rights legislation have been passed. With reference to the development from negative to positive rights legislation, negative and positive rights are sometimes referred to as the first and second pillars of linguistic rights.

With specific reference to language rights, negative rights amount to freedom from active discrimination on the basis of language. Positive language rights, on the other hand, are more extensive: these include rights such as the right to translation in legal contexts, right to access to public services in an appropriate language, and the right to the revitalisation of dying languages. Positive language rights, as tends to be the case of positive rights generally, typically involve large expenditure projects on the part of duty-holders. It is clear that positive in addition to negative language rights are necessary if the aims of the language rights movement are to be achieved: in many cases, minority languages will inevitably die out unless active investment is made in them and their communities.

The distinction between negative and positive rights may sometimes be problematic in that identifying positive rights as distinct from negative rights gives duty-holders a further argument against upholding certain rights. To deal with this problem, it may be argued that all rights are in fact positive, as even upholding negative rights may involve great investment on the behalf of states (such as investing in the development of effective law enforcement). Alternatively, it may be argued that all rights are both positive and negative: they all involve freedom from some form of oppression (whether it be passive discrimination or active violence) and all involve active expenditure on the part of the duty-holder.

Another philosophical point relevant to language rights concerns the special nature of language: language is a medium of interpersonal communication and so exists only in communities. It is at best unclear whether or not a language can be said to meaningfully said to exist for an individual. This presents particular problems given that the basic human rights paradigm conceives of rights as properties of individuals and not groups. Although, as mentioned above, some significant movement has been made towards recognising group rights, much of the existing human rights canon does conceive of rights as individual, and so it is within this well-developed paradigm that language rights have the most scope for development.

One final point to be considered is the universality of language rights. Language rights, like prototypical human rights, are theoretically universal in application. However, in practice language rights legislation has concentrated solely on minority groups. Despite the theoretical objections that clearly might be made to this, it makes pragmatic sense: it is minority groups that typically run the risk of having their language rights infringed. However, some writers have argued that it would be better to legislate for majority language rights also, with the aim of incentivising majority groups to aid in passing language rights legislation.