The US National Council on Radiation Protection and Measurement gave expression to the theoretical resolution of this human dilemma by articulating the implicit reasoning behind subsequent radiation protection standards development:[20]

1. A value judgment which reflects, as it were, a measure of psychological acceptability to an individual of bearing slightly more than a normal share of radiation-induced defective genes.
   
   
2. A value judgment representing society's acceptance of incremental damage to the population gene pool, when weighted by the total of occupationally exposed persons, or rather those of reproductive capacity as involved in Genetically Significant Dose calculation.
   
   
3. A value judgment derived from past experience of the somatic effects of occupational exposure, supplemented by such biomedical and biological experimentation and theory as has relevance.

This is now an internationally accepted approach to setting standards for toxic substances when no safe level of the substance exists.

        In short, this elaborate philosophy recognises the fact that there is no safe level of exposure to ionising radiation, and the search for quantifying such a safe level is in vain. A permissible level, based on a series of value judgments, must then be set. This is essentially a trade-off of health for some `benefit' -- the worker receives a livelihood, society receives the military `protection' and electrical power is generated. Efforts to implement these permissible standards would then logically include convincing the individual and society that the `permissible' health effects are acceptable. This has come to mean that the most undesirable health effects will be infrequent and in line with health effects caused by other socially acceptable industries. Frequently, however, the worker and/or public is given the impression that these `worst' health effects are the only individual health effects. A second implication of the standards-based-on-value-judgments approach is that unwanted scientific research resulting in public scrutiny of these value judgments must be avoided.
        The genetic effect considered by standard setters as most unacceptable is serious transmittable genetic disease in live-born offspring. These severely damaged children are usually a source of suffering for the family and an expense for society which must provide special institutions for the mentally and physically disabled. Severely handicapped people rarely have offspring; many die, are sterile or are institutionalised before they are able to bear children. Workers and the public are told that the probability of having such severely damaged offspring after radiation exposure within permissible levels is slight. By omission, a mildly damaged child or a miscarriage is implied to be `acceptable'.

        Standard setters judge that the most severe damage done directly to the person exposed is a fatal radiation-induced cancer, and again, this is a rare occurrence when exposure is within permissible levels. All other direct damage is by omission considered `acceptable'.
        In its 1959 report recommending occupational standards for internal radiation doses (i.e. radioactive chemicals which are permitted to enter the body through air, water, food or an open wound), the International Commission on Radiological Protection (ICRP) formed the following definition:

A permissible genetic dose [to sperm and ovum], is that dose [of ionising radiation], which if it were received yearly by each person from conception to the average age of childbearing [taken as 30 years], would result in an acceptable burden to the whole population.[16] [Emphasis added.]

        This might be paraphrased to say that the general public (governments) may be willing to accept the number of blind, deaf, congenitally deformed, mentally retarded and severely diseased children resulting from the permissible exposure level. Defined this way, the problem becomes primarily an economic one, since society needs to estimate the cost of providing services for the severely disabled. Once reduced to an economic problem, some nations may choose to promote early detection of foetal damage during pregnancy and induced abortion when serious handicap is suspected. When a foetus is aborted prior to sixteen weeks' gestation the event may not need to be reported and included in vital statistics. It becomes a non-happening, and the nation appears to be in `good health', having reduced the number of defective births.
        Mild mutations, such as asthma and allergies, are ordinarily not even counted as a `cost' of pollution. The economic burdens, `health costs', fall more on the individual and family than on the government. Their pain and grief do not appear in the risk/benefit equation. Parents and children are unaware of the `acceptable burden' philosophy.
        The prediction of the magnitude of the burden of severe genetic ills on an exposed population is essential to this philosophy. However, the data accumulated at Hiroshima and Nagasaki did not give clear answers. Either through ineptitude or loss of survivors of the bombing, who died before their story was told, the researchers failed to find any severe genetic ills clearly attributable to the parental exposure to radiation at low doses.[21] Probably the more fragile individuals in the population died from the blast, fire and trauma of the bombs, the women not surviving long enough to become pregnant.[22]
        Governments could not use the research on genetic damage in children of medical radiologists,[23] although this damage was measurable, because, in the early days, radiation exposure to physicians was not measured. No quantitative dose/response estimates could be derived.
        Animal studies of radiation-related genetic damage abounded, and the recommending body, ICRP, used (and still uses) mouse studies as a basis of its official predictions of the severe genetic effects of ionising radiation in humans.
        As late as 1980, a US National Academy of Science publication from its committee on the Biological Effects of Ionising Radiation[24] stated:

New data on induced, transmissible genetic damage expressed in first generation progeny of irradiated male mice now allow direct estimation of first generation consequences of gene mutations on humans . . . As with BEIR I, a major obstacle continues to be the almost complete absence of information on radiation-induced genetic effects in humans. Hence, we still rely almost exclusively on experimental data, to the extent possible from studies involving mammalian species [i.e. mice].

        These mouse studies are used as the basis of prediction, and permissible doses are set so that the expected number of severe transmittable genetic effects in children of those exposed could be presumed to be an acceptable burden for governments choosing a nuclear strategy.
        The introductory section of ICRP Publication 2, 1959, states:

The permissible dose for an individual is that dose, accumulated over a long period of time or resulting from a single exposure, which, in the light of present knowledge carries a negligible probability of severe somatic [damage to the individual] or genetic [damage to the offspring] injuries, furthermore, it is such a dose that any effects that ensue more frequently are limited to those of a minor nature that would not be considered unacceptable by the exposed individual and by competent medical authorities. Section 30.[16] [Emphasis added.]

        Mild mutations are notably happenings of a minor nature, normally neither reported nor monitored in the population. They are likely to be statistically hidden by normal biological variations and unconnected in the mind of the individual or his/her physician with the exposure. The publication continues:

The permissible doses can therefore be expected to produce effects [illnesses] that could be detectable only by statistical methods applied to large groups. Section 31.[16] [Emphasis added.]

        In spite of this clarity, no such statistical audit of all health effects including chronic diseases in exposed people and mild mutations in their offspring has ever been done. More than 25 years have expired since this document was published and the world is more than 35 years into the nuclear age.
        As late as 1965, ICRP Publication 9[25] stated:

The commission believes that this level [5 rems radiation exposure per 30 years for the general public] provides reasonable latitude for the expansion of atomic energy programs in the foreseeable future. It should be emphasised that the limit may not in fact represent a proper balance between possible harm and probable benefit because of the uncertainty in assessing the risks and benefits that would justify the exposure. [Emphasis added.]

        The committee protected itself against accusations of wrongdoing but failed to protect the public from its possible error. It defines its role as recommending, with the responsibility of action to protect worker and public health resting with individual national governments. Governments in turn tend to rely on ICRP recommendations as the best thought of internationally respected experts.
        In spite of this uncertainty about responsibility and safety levels for exposure of the public, 5 rem per year, rather than per 30 years, was permitted for workers in the nuclear industry. The 5 rem per 30 years was set as the average dose to a population, with a maximum of 0.5 rem per year (15 rem per 30 years) for any individual member of the public.
        For twenty years, between 1945 and 1965, health research on the effects of ionising radiation exposure has focused on estimating (not measuring) the number of excess radiation-induced fatal cancers and excess severe genetic diseases to be expected in a population (i.e. a whole country) given the average estimated exposure to radiation for the country. Disputes among scientists usually have to do with the magnitude of these numbers. Omitted from this research are other radiation-related human tragedies such as earlier occurrence of cancers which should have been deferred to old age or even might not have occurred at all because the individual would have died naturally before the tumour became life-threatening. These are not excess cancers, they are accelerated cancers. This approach also omits other physiological disorders such as malfunctioning thyroid glands, cardio-vascular diseases, rashes and allergies, inability to fight off contagious diseases, chronic respiratory diseases and mildly damaged or diseased offspring. The implications of such `mild' health effects on species survival seem to have either escaped the planners of military and energy technology, or to have been deliberately not articulated. Other obvious limitations of this national averaging approach include the failure to deal with global distribution of air and water with the result that deaths and the cumulative damage to future generations are not limited to one country.
        The usual procedure for setting the standard for a toxic substance or environmental hazard is to decide the relevant medical symptoms of toxicity and determine a dose level below which these symptoms do not occur in a normal healthy adult. This cut-off point is sometimes called the tolerance level and it represents a sort of guide to the human ability to compensate for the presence of the toxic substance and maintain normal health. The tolerance level for a substance, if one can be determined, is then divided by a factor (usually 10) to give a safe level. This allows for human variability with respect to the tolerance level and also for biological damage which may occur below the level at which there are visible signs of toxicity, i.e. sub-clinical toxicity.
        Human experience with ionising radiation had been recorded for more than fifty years prior to the nuclear age, the early history of handling radioactive material having been fraught with tragedy. The discoverer of the X-ray, W. K. Roentgen, died of bone cancer in 1923, and the two pioneers in its medical use, Madame Marie Curie and her daughter, Irene, both died of aplastic anaemia at ages 67 and 59 respectively. At that time, bone marrow studies were rarely done, and it was difficult, using blood alone, to distinguish aplastic anaemia from leukaemia. Both diseases are known to be radiation-related. Stories of early radiologists who had to have fingers or arms amputated abound. There were major epidemics among radiation workers, such as that among the women who painted the radium dials of watches to make them glow in the dark. Finally, there were the horrifying nuclear blasts in Hiroshima and Nagasaki.
        The painful period of growth in understanding the harmful effects of ionising radiation on the human body was marked by periodic lowering of the level of radiation exposures permitted to workers in radiation-related occupations. For example, permissible occupational exposure to ionising radiation in the United States was set at 52 roentgen (X-ray) per year in 1925,[26] 36 roentgen per year in 1934,[27] 15 rem per year in 1949[28] and 5 to 12 rem per year from 1959 (depending on average per year over age 18) to the present.[29] Recently there has been an effort to increase permissible doses of ionising radiation to certain organs such as thyroid and bone marrow[30] in spite of research showing the radiosensitivity of these tissues. This newer trend probably reflects economic rather than physiological pressures, especially given the lack of an acceptable audit of physiological cost.