I READ in this paper the other day that "we are now more honest about sex than about language".

This is embarrassing but true. It is quite obvious that our language relationships are kept secret, confessed in the dark only to census collectors and even then not entirely truthful. The fact is that we are not "comfortable" or "at home with our language, the way we are with our sexuality in the 1990s. We are lacking in linguistic confidence.

I do not mean that we are the illiterates of Europe (though it seems we are). But we have yet to build up a national army of professional language counsellors. There are no language problem pages in the magazines. We do not have language manuals or language therapy. We do not have The Good Language Guide, parts one and two, even in print, never mind available nationwide on video in full colour and impeccable taste.

We are as fearful and confused about language today as we were about sex in the Fifties (the decade, not the age group). Even those equally attracted to the Irish and English languages are ashamed of their bilingualism. "Coming out" as an Irish speaker is far from easy.

There is some ironic comfort in the fact that we are not we are alone in losing a language. In fact languages are being obliterated at a fierce and increasing rate. A research paper read last year to the American Association for the Advancement of Science reckoned that nearly 3,000 languages could become extinct over the next century.

However, Irish is not yet as threatened as Aakwo, Bella Coola, Grawadungalung, Santa and Zyrian or as Odut (a Palaeosiberian tongue, better known as Yukaghir) or as any of the Ge Pano Carib strain of languages, the most endangered of all.

One of the many Red Indian languages was recently buried in Massachusetts at the funeral of Red Thunder Cloud, a 76 year old storyteller and the last person to speak the tongue of the Catawba tribe. He left some recordings, and we know the words he would utter to his dog: Swie hay, tanty! (tr. Move, hound!"). The animal always seemed to understand.

This may be worth researching. Red Thunder Cloud's dog is still alive, and is thereby a repository of the Catawba language. If the Irish language is also to end up understood only by dogs, we will have to study their interaction. It is initially reassuring that Ireland's dog population is 22.7 per cent higher, in Gaeltacht areas, but a companionable grant aided dog for every Irish speaker should perhaps be our objective.

There is need for research too into exactly how Irish language speakers communicate, and how they recognise each other outside their clubs and other protected areas. It is as mysterious as the way in which Irish traditional musicians change tempo or choose to repeat a chorus at exactly the same time. The whole thing is tied up with sexuality and evolution and highly sophisticated body language (the original Esperanto).

Some of the answers to our Irish language problem may lie in the insect world and its use of song. Many cicadas, crickets, grasshoppers and other insects sing silently as a way of attracting the opposite sex and avoiding the attentions of predators - more or less the same twin aims of Irish speakers. Green lacewings and tremulating katydids, for example, jerk or circle their abdomens at 30-120Kz, shaking the substrate at the same time. Sensitive receptors in their legs are turned to the frequency range characteristic of their species.

We must also welcome minor genetic mutations in the Irish speaker species, particularly the singing variety(the majority, as it happens). When this occurs in the insect world, new species arise in a completely arbitrary way by both males and females choosing each other on the basis of novel songs.

Prof Charles Henry of the Department of Ecology and Evolutionary Biology, University of Connecticut, says this can happen if insect songs are under relatively simple genetic control, and with so many Irish songs sounding exactly the same, the theory has direct relevance. The professor says that a single mutation of the gene for singing could produce a new song unacceptable to the rest of the species, but "a single serendipitous success in mating by an individual bearing the mutation" would be all it would take to introduce the new gene to the population.

This may be the way forward. A lot of courage will be required, but then a lot is at stake.