This article is excerpted from the chief justice's commencement address at the George Washington University School of Law on Sunday :

No one questions that a criminal conviction should always be open to correct a miscarriage of justice. But no other system in the world invites our kind of never-ending warfare with society, long after criminal guilt has been established, beyond reasonable doubt, with all the safeguards of due process. Our system has moved thoughtful, sensitive observers who are dedicated to individual liberty to ask: "Is guilt irrelevant?"

Looking back, we see that over the past half century we have indulged in a certain amount of self-deception with euphemisms, sometimes perhaps to sugar-coat the acid pills of reality, and sometimes to express our humane aspirations for those who break our laws. Prisons became "penitentiaries" --pirations for those who break our laws. Prisons became "penitentiaries" places of penitence--juvenile prisons became "reform schools" and, more recently, we have begun "halfway houses" without being quite sure halfway from what to what.

I have long believed--and I have frequently said--that when society places a person behind walls and bars it has a moral obligation to take some steps to try to render him or her better equipped to return to a useful life as a member of society. Now, I say "try," and I use the term moral obligation, not legal, not constitutional. The Constitution properly mandates dues process and it mandates certain protective guarantees, but it mandates nothing concerning the subject of punishments except that they not be "cruel and unusal." To make these people good citizens is also for our own proper self-interest.

To do all the things that might have some chance of changing persons convicted of serious crimes will cost a great deal of money, and 1981, if you have listened to talk shows, is hardly the year in which to propose large public expenditures for new programs to change the physical plants and internal programs of penal institutions. So what I am about to propose are programs of relatively modest fiscal dimensions that I believe will help--but with no guaranteed results.

Two steps could reasonably be taken within the range of affordable expenditure. I relate them chiefly because they are affordable in an economic sense--and feasible in terms of the psychology and the political and economic realities of 1981. These proposals are closely related. Each bears on training and education--training of the inmates and training of the keepers.

In all too many state penal institutions, the personnel--the attendants and guards--are poorly trained and some are not trained at all for the sensitive role they should perform. A fairly recent study shows an astonishing rate of turnover of guards and correctional personnel.

An imporrtant and lasting consequence of lac of trained personnel is the impact on the inmate--the individual inmate--who continues his hostility toward society, toward fellow inmates and prison personnel. The keepers come to be the immediate symbols of the society that keeps them confined. Unfortunately, judicial holdings have not always discouraged this warfare. More often than not, and I emphasize this, inmates go back into society worse for their confinement.

The best of prison administrators cannot change some of the negative conditions unless those in the high-turnover, lower-echelon categories are carefully screened, well-trained and reasona!ly paid. Psychological testing of applicants is imperative to screen out people with latent tendencies of hostility. The existing prohibitions on psychological screening must be reexamined. Today, those lower positions in most of the states are generally not paid adequately enough to get even minimally qualified people.

The time is ripe to extend the fine work begun in 1972 by the National Institute of Corrections, and we should proceed at once to create a national academy of corrections to train personnel much as the FBI has trained local and state police. This is especially needed for the states, which have no real training resources available. The academy should also provide technical assistance to state and local institutions.

The second step for which I would urge consideration is one that would need to be over a longer period of time. We should introduce or expand two kinds of educational programs in the prisons.

The first would be to make certain that every inmate who cannot read, write, spell and do simple arithmetic would be given that training--not as an optional matter but as a mandatory requirement. The number of young, functional illiterates in our institutions is appalling. Without these basic skills, what chance does any person have of securing a gainful occupation when that person is released and begins the search for employment --with the built-in handicap of a criminal conviction?

Focusing on the longer-term prisoner, the second phase of this educational program would require a large expansion of vocational training in the skilled and semi-skilled crafts. The objective would be that a prisoner would not leave the institution without some qualifications for employment in the construction, manufacturing or service industries. These vocational training programs should also be mandatory. An inmate who declines to cooperate must be motivated to do so by incentives including shortening the sentence. Just as good behavior credit is now allowed to reduce sentences, we should allow credit on sentences for those who cooperate by learning. We should help them to learn their way out of prison.

Even in this day of necessary budget austerity, I hope that the president and Congress, in whose hands such matters must rest, will be willing to consider these two modest, but important steps. No one can guarantee results, but if we accept the moral proposition that we are our brothers' keepers and the moral proposition that there is a divine spark in every human being--hard as that is to believe sometimes--we must try.