First, the obligatory declaration: both my parents left school before their 14th birthdays, one grandfather drove trucks and the other was a shot-firer on the South Yorkshire coal field. Actually, that makes one of them a member of the labour aristocracy, or in the eyes of hardline members of the National Union of Mineworkers a class traitor. And indeed, the windows of his house were smashed during the First World War when the miners were on strike and the deputies and shot-firers kept the pits open so that the miners would have somewhere to work when the strike was over.

There is a decent working-class tradition behind the behaviour of those semi-striking academics who last year were willing to withhold examination results from their employers but not willing simply to abandon the marking process entirely.

Back to the children of the uncertificated. It has been interesting to watch the argument between those who really, even if secretly and in the privacy of their own hearts, want the playing field not so much levelled as tilted against the fortunate and those who rally to the cry of "no social engineering here".

Start at the beginning. It is certainly true that young people whose parents attended university are themselves much more likely to apply to university than those whose parents did not. Even when you strip out the double-whammy that those whose parents attended university come from - generally - better-off homes than those whose parents did not, there's still a marked effect.

³ÉÈËVRÊÓƵ

Part of the reason for this emerged during a survey conducted a decade ago by the National Foundation for Educational Research on behalf of the ever-anxious Oxford University admissions office. Asked where they got their advice about higher education, pupils listed friends, parents, teachers and prospectuses - in that order. This was not uniform, but it did make those of us who spend hours constructing prospectuses wince a bit. If it is still true that friends and parents are the first source of advice, it is highly likely that friends whose parents went to university will give more informed advice and will be more encouraging about the process than those whose parents did not, and certainly true that university-educated parents will give more informed advice and will have more ideas about what sort of information they don't have but might get.

But that interesting phenomenon is irrelevant to the Universities and Colleges Admissions Service's bright idea that admissions tutors should know whether applicants' parents went to university. For the people Ucas knows about are precisely those for whom the lack of a university-educated parent has clearly not been inhibiting - they have applied. The hurdle has been cleared. So, why would it benefit an admissions tutor to know whether an applicant has university-educated parents? Presumably, there is some empirical evidence that a first-generation student outperforms a second-generation student who arrives with the same A levels and GCSEs. I say "presumably" because I've not seen it.

³ÉÈËVRÊÓƵ

It is widely believed that a state-school student will outperform a private-school pupil who arrives with the same A levels and GCSEs; and it seems to be thought that the same phenomenon applies with first-generation students. But we may be relying on antiquated evidence. In the good old days, state-school boys beat private-school boys, who beat private-school girls, who beat state-school girls - even when you cleaned up the data to allow for the fact that different disciplines produced very different results and had very different sex and class ratios. That no longer seems to be true, although current data is hard to gather except at inordinate cost in time and money. And the higher education system has become so non-uniform that generalisations across the entire sector are extremely hazardous.

In fact, the point at which the information would be useful is not at admissions time, though it might be marginally helpful if one couldn't work out why an apparently intelligent and well-organised applicant seemed to be applying for unsuitable courses. Even then, one's first port of call would surely be the school - it signs off on the form, and ought to do so in not too perfunctory a fashion. No, the point at which one really wants to know everything about a student is when he or she arrives to start a course.

It's then that the question of how much they take for granted, how much is deeply mysterious to them, and how hard they will find it to cope is really raised. Being a first-generation student is even then only one small part of the several thousand-piece jigsaw puzzle that makes up each and every student.

Alan Ryan is warden of New College, Oxford University.

³ÉÈËVRÊÓƵ