Monday, January 19, 2004



Chapter 46 from: J.J. Ray (Ed.) "Conservatism as Heresy". Sydney: A.N.Z. Book Co., 1974

ARE RACISTS ETHNOCENTRIC?



ALTHOUGH COINED much earlier, the word 'ethnocentrism' was first popularised in the epochal work of Adorno, Frenkel-Brunswik, Levinson & Sanford (1950). These authors used the term in preference to 'racial prejudice' on the grounds that not all prejudice they were concerned with was racial (e.g. anti-Semitism) and not all was negative (i.e. one can be prejudiced towards as well as against). Generally, then, 'ethnocentrism' in their terms means something like 'thinking of one's own ethnic group as providing a standard from which other groups diverge only to their discredit'

It must be stressed that, although disguised as a definition, what we have here is in fact an explanation of racism. It is asserted implicitly that at least some racists dislike other ethnic groups because they have an excessively high regard for their own group. It is assumed that thinking well of one's own group has as its mirror-image thinking ill of other groups. In two-value logic, this is an impeccable assumption. The trouble is that we do not live in a two-value world. In fact, nowhere has this been better demonstrated than in the follow-up work to Adorno et al, (1950). Early attempts to reverse the wording of the 'F' scale (summarised in Brown) 1965), did assume a two-value world. They thought that if a person agreed with an original item he should disagree with its contradictory. Empirically, it was found that this did not happen. The solution given by Christie, Havel & Seidenberg, (1956) to the conundrum that this phenomenon posed was that psychological opposites were not the same as logical opposites.

It would appear then, that for all their logical opposition, racism (thinking ill of other groups) and ethnocentrism (thinking well of one's own group) might not in fact be highly correlated. Only an empirical enquiry can find out. What Adorno et al, (1950) implicitly assumed needs testing.

That the assumption could be false becomes more apparent when we consider alternative hypotheses for the explanation of racism. It could, for instance, be the case that ethnocentrism is generally merely one aspect of a general trait of misanthropy. People who don't like other races may not like anyone very much. In this case we would expect a positive correlation between ethnocentrism and racism rather than the negative correlation implicitly postulated by Adorno et al, (1950). Thinking ill of other groups may go with thinking ill of one's own group also.

Going beyond this, the most cautious hypothesis of all might be that no generalisation is possible. Attitudes to various groups might be stimulus-specific. We like groups who have features or characteristics that we have been conditioned to like and that is all there is to it. Just because we don't like Jews, it does not mean we will not like Aborigines or that we will like our own society. This hypothesis does at least have to recommend it the fact that the things objected to in one group are often not the same as the things objected to in other groups. Blacks are said to, be 'dirty' while Jews are said to be 'money-hungry'. Presumably one could approve of one attribute while not approving of the other. If so, one would dislike the one group but not the other. Prejudice would be specific -- not general. This raises the possibility that prejudice could in some sense and for some individuals be 'rational'. If one has been thoroughly brought up and, conditioned to place a high value on hygiene and one particular group is in fact characterised by carelessness of hygiene, it is surely in no sense irrational to dislike members of that group. This explanation for prejudice is at least a somewhat simpler one than the complex psychodynamic one proposed by Adorno et al, (1950). It may be that a punitive father produces prejudiced children but the connection is somewhat less direct than saying that disapproval of unhygienic behaviour generalises to disapproval of any particular group of people characterised by such behaviour. Indeed, when a mother socialises her child with words such as 'you dirty little boy' it is as much the person she is condemning as the behaviour. In conditioning an abhorrence of dirt, she is also conditioning an abhorrence of dirty people.

Since it is highly likely that care for hygiene is adaptive, it could be that any group which is not hygienic in some respect must attract 'prejudice' from members of our society. The prejudice could be an undesirable side-effect of something that is of itself inherently desirable.

In one sense, then, there are two polar hypotheses for the explanation of prejudice: the 'rational' and the 'irrational'. The first says that prejudice is not prejudice at all, but rather postjudice. One does not dislike blacks because one has some deep Freudian need to do so, but rather because one has come in contact with blacks and found out as an empirical fact that they are characterised by attributes which one personally does not approve of in anyone -- be they blacks, Jews, Tibetans or one's own brother.

The second hypothesis, on the other hand, is the one with which social scientists are more familiar -- that prejudice is 'irrational' in the sense of springing from deep personal inadequacy. One's own inadequacy is so great and so disturbing that it must be masked by stigmatising other weaker groups as yet more inadequate. It will be seen then, that this paper challenges very extensively any notion of generality in the phenomena of prejudice. Not only does it challenge the view that thinking well of one's own group implies thinking ill of others, but it also challenges the view that thinking ill of one out-group implies thinking ill of another outgroup. The first of these assumptions has passed almost unnoticed in literature on the topic, (for one exception, see Hoogvelt, 1969) but the latter has been the subject of some test.

In fact there has been a consistent record of confirmation for this second hypothesis. There is always some correlation between various forms of anti-out-group feeling. Adorno et al, (1950) found a significant correlation between their anti-Semitism and 'ethnocentrism' scales. In fact, at .80, the correlation was not only significant but high enough to suggest practical identity between the entities concerned. The so-called 'ethnocentrism' scale used in these authors' study, however, was not an ethnocentrism scale in the narrow sense suggested here (thinking well of one's own group) but a rather broad-ranging prejudice scale falling into three subscales labelled respectively 'Negroes', 'Minorities' and 'Patriotism'. Of these three, patriotism might seem closest to what is here called ethnocentrism but even so, it includes items dealing with 'Germans and Japs', 'Mexico', the inevitability of war and so on. The correlation between these three sub-scales was also high-ranging from .74 to .83. The lack of thematic unity in the sub-scales must, however, make one a little suspicious of these correlations. Why, for instance, was the item dealing with 'Germans and Japs' not placed in the 'Minorities' sub-scale? Since this item and others like it in fact went into the 'Patriotism' sub-scale, it must come as no surprise that the 'Patriotism' and 'Minorities' sub-scales correlated so highly.

For all that, it must be conceded that there is at least some generality in all anti-out-group attitudes. Ray & Doratis (1972) reported in fact that even prejudice against religious out-groups correlated with ethnic prejudice. This recurrent correlation must be taken as some support for the 'irrational' explanation of racial prejudice. If prejudice is completely general, it cannot be stimulus-specific or stimulus-bound unless the objects of prejudice are all characterised by the same properties -- and any such set of common properties would be extraordinarily hard to find.

It must be observed, however, that just how general prejudice is, is not really known. Almost all ethnic prejudice scales assume it to be general and hence mix up in the one scale items dealing with a variety of ethnic groups. This was certainly so of the Adorno et al, (1950 ) 'ethnocentrism' scale. It would be much more informative if such a priori assumptions were not made and separate scales used to assess attitude towards each particular ethnic group. If, after so doing, it were found that the separate scales correlated very highly, the hypothesis that prejudice can be stimulus-specific would have to be abandoned. If, on the other hand, it were found that the correlation were only modest, one would have to say that there were some stimulus-specific elements in ethnic prejudice and some general elements. Putting it another way, one could say that for some people prejudice is the result of 'rational' processes and for some others it is the result of 'irrational' processes.

If we assume that 'rational' (stimulus-bound) prejudice is possible, another possibility that arises is that one ethnic group might attract it more than another. Prejudice towards aborigines, for instance, might be almost entirely 'rational' whereas prejudice towards Jews might be almost entirely 'irrational'. There might be different underlying causes for prejudice towards different groups. This, then, constitutes another reason why prejudice towards different groups ought to be treated differently and not lumped together to form a single 'racial prejudice' scale.

The experimental hypotheses so far, then, are: 1. That ethnocentrism will not be significantly correlated with any form of racism; 2. That different forms of racism will be significantly correlated but that the correlations observed will fall far short of suggesting identity.

One difficulty with the second hypothesis is the difficulty of knowing what would constitute sufficient evidence to reject it. How near is 'far short'? Obviously, there is little assistance to be gleaned here from statistical criteria. Underlying the second hypothesis is the view that 'rational' prejudice is possible. The second hypothesis is merely one possible deduction from that underlying theory. To strengthen the test of the theory, therefore, it seems desirable to derive yet a further deduction to form a third hypothesis. If two deductions from the theory are confirmed, the theory will be much more effectively supported.

The second deduction to be made (Hypothesis 3 ) depends on the fact that in Australia, prejudice against Aborigines (blacks) appears to be positively correlated with contact frequency. Overall, Australians at a referendum of a few years ago showed that they were overwhelmingly in favour of increased government action to assist Aborigines. In such circumstances, news reports of anti-Aborigine prejudice are widely noted and are responded to with some outrage. Such reports, however, invariably come from country towns which have an Aboriginal settlement nearby or from depressed suburbs of big cities where Aborigines congregate. We might infer then that most Australians have a rather unrealistic view of Aborigines -- regarding them as 'just the same as us only with a brown skin'. The cultural differences inevitable in a people who were in the stone age only a few generations ago do not appear to be anticipated. When however white Australians are brought face to face with Aborigines, the vast cultural differences soon become apparent and prejudice is evoked where none was before.

The second deduction, then, is that the correlations between the various forms of racism will be less in geographical areas where there is high contact with the out-groups concerned. The rationale for this deduction is as follows: If prejudice is normally composed of both rational and irrational elements, only the rational elements can be affected by contact. Whether contact with the group which one dislikes leads to an increase or a decrease in rational dislike will be a function of the stimulus properties of the target-group concerned. But since, by definition, irrational prejudice is a function of the prejudiced person's deep needs rather than of the stimulus characteristics of the target group, contact with that target group will be irrelevant to that irrational prejudice. Irrational prejudice can only be changed if the underlying needs which caused it change. Irrational prejudice then can exist prior to any contact with the groups concerned and will not be changed by such contact. Prior to contact, in fact, all prejudice must be irrational prejudice. Therefore, in people who have had little contact with Aborigines in particular, prejudice should be in absolute terms less prevalent and at the same time highly correlated. Since any target group will do for irrational prejudice, there should be no need to discriminate between such groups. If you are prejudiced against one, you will be prejudiced against the other.

After contact with the group or groups concerned, however, this relatively cosy little picture must change somewhat. If, for instance, people suddenly find themselves living in close contact with Aborigines and Aborigines happen to be in fact rather unhygienic in their habits, some people previously without prejudice will start to say that they don't like Aborigines. But the same people will still have no reason to dislike Jews. The two forms of prejudice will not correlate. The addition of a few rationally `prejudiced' persons to the ranks of the prejudiced (previously composed only of irrationally prejudiced persons) should attenuate or dilute the previous high correlation between the various forms of prejudice. In operational terms, then, the third hypothesis for this study falls into two parts: that anti-Aboriginal prejudice will be highest in suburbs where large numbers of Aborigines live, and that in such suburbs, the correlation between anti-Aborigine sentiment and racial dislike of other forms will be lower than in suburbs where there are few or no Aborigines.

METHOD

This study took advantage of the fact that, at the time it was carried out, one particular suburb of Sydney, Redfern, was the centre of a national controversy because of its citizens' and local council's desire to prevent the formation of an Aboriginal housing co-operative in the area. Funds for the project were granted by the Federal government and the objections of local residents were overruled. Redfern already in fact had what was probably one of the largest urban concentrations of Aborigines in Australia. It was hence ideal for the present study.

To provide a standard against which attitudes in Redfern could be compared, however, it was necessary to find another suburb of equivalent socio-economic status which did not have an Aborigine population. The only suitable suburb was Zetland. Both Redfern and Zetland were rated 7 (the lowest rating) on Congalton's, (1969), scale of prestige of Sydney suburbs. Although only Redfern had a large population of Aborigines, both Redfern and Zetland had a large population of recent Southern European and middle Eastern migrants. Neither suburb appeared to have any detectable Jewish population.

For the purposes of the study, a questionnaire was designed which contained four separate scales to measure attitude towards the four ethnic groups which were thought to be most relevant to a study of prejudice in Australia: Jews, Aborigines, Southern European migrants (Italians, Greeks etc. ) and Australians themselves. The latter scale was intended as the measure of ethnocentrism among Australia's Anglo-Saxon majority. All four scales had ten items each -- five positively worded and five negatively worded.

It was perhaps regrettable that all four scales had to be new constructions -- no already published prejudice scales were suitable for use on this occasion. This was inevitable, because of the phenomenon mentioned earlier; that most previous work appears to have assumed that the different forms of racial or ethnic dislike invariably go together. Previous scales therefore are in general a jumble of items referring to a whole variety of out-groups. Nonetheless, individual items of previous scales did of course form the cores around which the new battery was built.

The four scales were all of equal length to facilitate comparison of means, reliabilities and other summary statistics. The items comprising the four scales were presented in the questionnaire interspersed among one-another. All were scored so that a high score represented a more 'ethnocentric' or prejudiced response.

Sampling was carried out by a small class of Sociology students as part of their course work. Each student took up to six questionnaires and went from house to house in the suburbs concerned. Each student selected his starting point from a map of the suburb with the aid of a pin and blindfold. He then visited houses in the block which was nearest to where the pin fell. He visited as many houses as was necessary to obtain three filled-out questionnaires from each suburb. Some questionnaires were filled out while the students waited and others were filled out at the householder's leisure and picked up later. The questionnaire was designed to be totally self-administering in order that some approach to standardised conditions of administration could be claimed. All questionnaires were checked for genuineness (to see that the students themselves had not filled them out) and no major problems were encountered: To safeguard confidentiality, the questionnaire was sealed in an envelope immediately the student received it back from the householder and the questionnaire was in any case anonymous. The location at which the person was contacted was however later written on the envelope by the student. This enabled data drawn from Redfern and data drawn from Zetland to be kept separate.

Because the conceptual focus of the study was the attitude of Australia's Anglo-Saxon majority towards the various minorities resident in Australia, the sampling necessarily excluded members of those minorities. If one of the student-interviewers encountered a coloured or European person when they went from door to door, various artifices were used by the students to extract themselves from the situation. If this did not prove possible, the interview was carried out but the results were not used.

RESULTS

A total of 178 questionnaires were actually completed -- eighty-eight from Redfern and ninety from Zetland. There were no significant differences between the two samples on the five demographic variables of Sex, Age, Education, Occupational Status and Political party preference. Item analysis based on the combined samples showed that all scales used worked reasonably well-with reliabilities (coefficient 'alpha') of .78, .79, .81 and .64 being recorded for the attitude to Aborigines, the attitude to Jews, the attitude to Southern Europeans, and attitude to Australia scales respectively. Considering the brevity of the scales and the fact that no pretesting had been carried out, the reliabilities of the first three scales were in fact quite high. All items used correlated significantly with their respective scale totals -- even after correction for overlap.

Contrary to the hypothesis, attitude to Aborigines was not significantly less favourable in Redfern than in Zetland. Means and SDs on the scale were 28.55 (7.05) and 28.63 (7.03) respectively. As the expected corollary of such a finding, the correlations between the various prejudice scales also differed little between the two samples. It had been expected that the correlation between attitude to Aborigines and attitude to Jews would be less in Redfern than in Zetland but the observed difference .333 versus .327 was not significant. The correlation between attitude to Aborigines and attitude to Southern Europeans was .368 in Redfern and .202 in Zetland. This was in the opposite direction to that expected but again was not significant. There was, then no support for either part of hypothesis 3.

Even the above results, however, do reveal that the second hypothesis was strongly supported. The correlation between different forms of racial dislike was in some cases extremely low. Taking such correlations across the entire group (both samples pooled), the attitude to Aborigines scale and the attitude to Jews scale correlated .330 while the attitude to Aborigines and the attitude to Southern European scales correlated .291. The attitude to Jews scale and attitude to Southern Europeans scale, however, correlated quite highly -- at .567.

Perhaps even more strongly supported, however, was the first hypothesis: one's attitude to Australia, Australians and things Australian correlated very little with one's racial dislikes -- .050 with attitude to Aborigines, .039 with attitude to Jews and .216 with attitude to southern Europeans.

DISCUSSION

It must be remembered that this study is principally the study of an assumption -- the assumption that ethnic likes and dislikes are unidimensional. Such an assumption occurs on every occasion when a single scale purporting to measure 'racial prejudice' or the like is used (e.g. Beswick & Hills, 1969). As has been mentioned, it is an assumption which has flowed at least in part from the early work by Adorno et al, (1950) which demonstrated extremely high correlations between measures of various forms of ethnic prejudice -- correlations so high that the assumption of uni-factoriality seemed justified. It has in fact almost become conventional wisdom that the man who is prejudiced in one respect will also be prejudiced in others -- the man who dislikes blacks will not like Jews either, and so on.

It is felt that the present results, although not widely generalisable because of sample limitations, do definitely show that there are at least some occasions when the assumption does not hold. In particular, the present results have given grounds for believing that among many Australians, attitude to Aborigines stands on its own -- whether you like or dislike Aborigines will not be very predictive of your attitude to other potential out-groups. The correlations observed do of course show that anti-Aboriginal people are slightly more likely to be prejudiced towards other groups (eight to eleven per cent shared variance) but, at the same time, the correlations are low enough to allow for there being large numbers of people who dislike Aborigines only. If by 'racist' we mean someone who is prejudiced against practically all out-groups, we can say from the present results that disliking Aborigines does not necessarily make you a 'racist'.

Attitude towards Southern Europeans and attitude towards Jews, however, do fit the classical picture of being highly correlated. How may this, then, be interpreted? It certainly cannot well be interpreted by the classical Freudian model; which sees prejudice as essentially irrational and more a reflection of the prejudiced person's deep projective needs, than of the disliked groups' stimulus properties. If it were all so irrational and projective, why would it matter what the out-group involved was? Obviously, it does matter. Attitudes towards the various groups are clearly differentiated.

The model which sees prejudice as at least partly rational, however, has no difficulty in fitting the results observed here. The different groups have different stimulus properties, so different people will dislike them. Jews are very different from Aborigines. If we can say that Aborigines are 'down and out', then we must say that Jews are 'up and out'. People who don't like sloth will not object to Jews, but they may object to Aborigines. People who do not like grasping materialism, will certainly find no fault with Aborigines but they may find fault with Jews. The fact that attitude to Jews and attitude to Southern Europeans are highly correlated would, on this account, simply show that the two groups have similar disliked properties -- such as being 'money-hungry' and often speaking with a heavy accent.

An intimate part of the Freudian 'projected-hostility' model of race prejudice, put forward by Adorno et al, (19,50) was of course that exaggerated dislike of out-groups was matched by exaggerated loyalty towards the in-group -- i.e. that positive ethnic sentiment ('ethnocentrism') would be highly related to negative ethnic sentiment ('racial prejudice'). The very fact that we have the word 'racism' in our vocabulary embodies an assumption that the direction of the racial sentiment does not need to be stated. It embodies the assumption that if you are 'prejudiced against' (one group) you will also be 'prejudiced towards' (your own group). If the term 'racial prejudice' embodies the assumption that people who dislike one out-group will tend to dislike all out-groups, then the term racism embodies the even broader assumption that race as such will be an obsession to people who show race-polarised dislikes and that such people will have a systematic belief in the superiority of one race or ethnic group -- a superiority from which the inferiority of all other groups must follow.

It is precisely this pervasive assumption that the present work has most clearly upset. This study would appear to have been the first which quite clearly allowed people to show positive ethnic sentiment ('ethnocentrism') without at the same time being committed to a particular stance on negative ethnic sentiment ('racial prejudice'). Because the two types of attitude were measured and reported separately it became possible for phenomena such as general misanthropy (disliking both the 'in' and the 'out' group) and general benevolence (liking both the 'in' and the 'out' groups) to emerge. Most importantly of all, however, it allowed for the possibility of there being no generality in ethnic attitudes at all.

It was then shown that people's opinion of their own group ('Australians') had absolutely no bearing on their attitude towards Jews or their attitude towards Aborigines. Some super-patriots were racially tolerant and some people totally alienated from Australian society were racially prejudiced. All possible combinations were equally likely. Only in the case of attitude towards Southern Europeans was there some relationship with Australian ethnocentrism. Although here the relationship was so weak (five per cent common variance), that some of the possible combinations of sentiment became only slightly less likely. These results also, then, represent a failure to confirm the expectations derived from the Freudian model of Adorno et al, (1950). Again however, there is no trouble in fitting the results to the 'rational' model of ethnic sentiment. If positive and negative sentiment do not have a common underlying psychodynamic cause, but are instead simply a function of the varying stimulus properties of the groups concerned, there is in fact no reason for the two types of sentiment to go together unless the stimulus properties of the two groups are indistinguishable. That they do not go together simply shows that the various groups are distinguishable and that the parameters on which they differ are important ones in the value-systems of the people contained in the sample.

Given the fact that the two generally important hypotheses of the present work came out in a direction which is best fitted by the 'rational prejudice' model, it is of course disappointing that the hypothesis of more local or topical importance was not confirmed. Redfern and Zetland did not differ in degree of anti-Aborigine sentiment. This obviously could be the outcome of several factors. It could of course mean that the model from which the hypothesis was deduced was inaccurate, but it is perhaps more likely that some of the necessary local or topical premises which were also involved in deriving the hypothesis, were not in fact accurate. It could for instance be the case that the people contacted in Redfern did not in fact have a history of greater contact with Aborigines than did the people contacted in Zetland. Regrettably, there was nothing in the questionnaire which allowed the extent of contact for each individual to be checked. Such a refinement should figure in any future studies similar to the present one. The reason why Redfern-dwellers were expected to have had greater contact was simply that Redfern is the Aboriginal enclave of Sydney. Geographically, however, Redfern is a rather elongated suburb and Aborigines are very clearly clustered towards its western end. Unfortunately, an informal check with the student-interviewers suggested that most of the sampling occurred in the central area rather than at the extremes. This, then, would have effectively invalidated one of the assumptions upon which the third hypothesis was built.

We may, then, conveniently summarise the findings of the present study by saying that, at least for the samples studied, the three assumptions embodied in the terms 'racial prejudice' and 'racism' were shown to be false. These assumptions are: 1. That positive prejudice (ethnocentrism) goes together with negative prejudice (dislike of other ethnic groups); 2. That dislike of one ethnic outgroup automatically goes with dislike of all ethnic out-groups; 3. That dislike of a particular ethnic group is necessarily prejudice. It could equally be postjudice -- i.e. a dislike stemming from experience with the group concerned, from experience of its members' real stimulus properties. It need not be a sentiment imported into the interracial contact situation which dictates from the beginning the formation of negative perceptions. Since both the above terms have, then, been shown to embody defective assumptions, a sounder replacement term might be 'ethnic sentiment'.

A remaining objection at this stage might be: 'But prejudice is something that happens at the individual level: even if it should be true that Aborigines are in general dirty or lazy, that still does not mean that any individual Aborigine will be dirty or lazy. By 'prejudice' one does not mean pre-judging the group. One means pre-judging the individual on the basis of what the group is supposed to be like. What is wrong basically is this stupid business of categorising people'.

One might remark that if it is wrong to categorise people as Aborigines, then it is surely just as wrong to categorise them as 'racists'. If it is categorisation of itself that is wrong then there would be few nouns left in the language for us to use. Fortunately or unfortunately, we do invariably work with general categories of phenomena. Take the noun 'dog'. There is in fact no such thing as a dog. There are only particular dogs -- animals that vary in size, shape, colour, smell, sound and practically everything else. Yet from this great variety we do extract a generalisation or abstraction and call it a 'dog' -- even though there is in fact no such thing as a typical dog. Viewed objectively, the whole phenomenon seems to be really peculiar behaviour -- an unparalleled exercise in vagueness. Why do we do it? For the simple reason that communication would be impossible if we did not. If we had a separate word for each individual dog and no general term, a lifetime would not suffice for us all to acquire the vocabulary necessary to discuss dogs. So it is not categorisation or labelling people' that is wrong or stupid. We label everything. We must.

What is wrong is perhaps then forming false expectations of the individuals so labelled. Can this however be helped? If I see that a dog is an Alsatian, I will avoid it because I have heard that Alsatians are often savage. That is probably adaptive behaviour for me -- even though I might have been doing that particular Alsatian a great injustice. It might in fact have been the most gentle and loving dog in the world. It might have been an exceptional Alsatian. If I happen to see that particular Alsatian often, I might quite soon learn that it is not in fact savage and I might in fact become good friends with it. This however does not mean that my initial behaviour was stupid or wrong. It simply means that never in life do we have all the relevant information about a particular situation in which we need to make some decision. We always, then, are thrown back on the need to rely on generalisations. This will mean that in some cases we will definitely be wrong and in consequence make bad decisions. It is however better to make some wrong decisions than to be permanently paralysed in indecision. If we never acted until all the information were in, we would spend life permanently rooted to the one spot. It is clear then, that every expectation, every generalisation, is in one sense prejudice. It is a prejudgment of the outcome of some action or situation based on imperfect information. In this sense prejudice cannot be right or wrong. It is simply unavoidable.

In interracial contact situations, therefore, the only cure for prejudice is to insure that other characteristics are not correlated with ethnic origin. While they are, people will use them, consciously or unconsciously, to form generalisations. For groups such as Australia's Aborigines, this means that a concerted effort must be made to raise their living and educational standard to the same level as that of white Australians. Calling people 'racists' and categorising them as evil, stupid or sick will probably be wrong and will certainly solve nothing.

REFERENCES

Adorno,T.W., Frenkel-Brunswik, E., Levinson, D.J. & Sanford, R.N. (1950). The authoritarian personality. New York: Harper.

Beswick, D.C. & Hills, M.D. (1969) An Australian ethnocentrism scale. Australian J. Psychol. 21, 211-226.

Brown, R.(1965) Social psychology. N.Y.: Free Press.

Christie, R., Havel, J. & Seidenberg, B.(1956) Is the 'F' scale irreversible? J. Abnorm. Soc. Psychol. 56, 141-158.

Hoogvelt, A.M.M. (1969) Ethnocentrism, authoritarianism and Powellism. Race, 11, 1-12.

Ray, J.J. & Doratis, D. (1972) Religiocentrism and ethnocentrism: Catholic and Protestant in Australian schools. Sociological Analysis 32, 170-179.



APPENDIX I

The attitude to Aborigines scale. Five-point response option to each item from Strongly Agree (5) to Strongly Disagree (1). Items marked 'R' are reverse scored.

1 Aborigines generally don't show much inclination to work.
2 Aborigines have been unfairly discriminated against. R
3 Aborigines are not very hygiene-conscious.
4 Drunkenness is one of the greatest problems with Aborigines.
5 Aborigines often get into fights with one-another.
6 Given the chance, the Aborigine will work as hard as the white man. R
7 It is only because they haven't had the same chance to get an education that Aborigines can't get work. R
8 Aborigines are a kind and gentle people. R
9 The Aborigines are a rather ugly race.
10 We could learn a lot from the way Aborigines always share with one-another everything they've got. R


APPENDIX II

The attitude to Jews scale. Five point response option. Items marked 'R' are reverse scored.

1 Jews tend to be very money-hungry.
2 Jews are as honest and public spirited as anyone else. R
3 A major fault of the Jews is their conceit, overbearing pride and their idea that they are a chosen race.
4 If it weren't for the Jews, we wouldn't have had many of our most brilliant scientists and thinkers. R
5 Jews are definitely not to be trusted.
6 The Jews are much to be admired for having kept their identity in spite of centuries of persecution. R
7 Jews are one of the most cultured races on earth. R
8 After what they suffered under Hitler, Jews deserve the support of people everywhere. R
9 It can't be an accident that so many people over the centuries have wanted to suppress the Jews; they must have done something to deserve it.
10 One trouble with Jewish businessmen is that they stick together and prevent other people from having a fair chance in competition.


APPENDIX III

Attitude to Southern European migrants scale. Five point response option. Items marked. 'R' are reverse-scored.

1 European migrants should make more of an effort to learn English.
2 Italians and Greeks work such long hours that they tear down the conditions won by the Australian workers.
3 European migrants have benefited us by increasing the range of foodstuffs we can buy. R
4 Greeks just don't know how to return the friendliness of the Australian people.
5 Italians are no better or no worse than any other people R
6 Migrants have done a lot to build up Australia. R
7 Italian and Greek migrants have lowered the standard of living in Australia because they all crowd together in small houses and live like animals.
8 I can't imagine any member of my family marrying a Greek.
9 Australia is a much more cultured place because of the many southern European migrants we have received. R
10 One thing you have got to say for migrants is that they are hard workers. R


APPENDIX IV

Attitude to Australia and Australians scale. Five-point response option. Items marked 'R' are reverse-scored.

1 Australia is a dump of a country. R
2 Anyone who wants to get on has to go overseas rather than stay in Australia. R
3 People who knock Australia don't know what they're talking about.
4 Australia always seems to be at least five years behind anywhere else. R
5 Australia has one of the best climates in the world.
6 Australia must be a pretty good place or all the migrants wouldn't want to come here.
7 The Australian way of doing things is pretty hard to beat.
8 Not many other countries have the advantages Australia has.
9 A basic trouble with a lot of Australians is a lack of imagination and stick in the mud attitudes. R
10 Australians are far too easy going for their own good. R

POST-PUBLICATION ADDENDUM

The most recent versions of the racial attitude scales are given here. See also here for a later application of the Australian patriotism scale. Follow-up research to the research reported above can be found here.