HomePersonalDemocratic PeaceDemocide20th C. DemocideMegamurderersLesser MurderersWhy DemocideDimensionsConflictMethodsTheoryPolicyLinks

Volume 3

Expanded Contents


1. Introduction and Summary
2. Aggression and the Conflict Helix
3. Frustration, Deprivation, Aggression, and the Conflict Helix
4. Misperception, Cognitive Dissonance, Righteousness, and Conflict
5. Marxism, Class Conflict, and the Conflict Helix
7. Cross-Pressures, Overpopulation, Anomie, and Conflict
8. Conflict as a Process and the Conflict Helix
9. Opposition, Determinism, Inevitability, and Conflict
10. Intentional Humanism

Other Volumes

Vol. 1: The Dynamic Psychological Field
Vol. 2: The Conflict Helix
Vol. 4: War, Power, Peace
Vol. 5: The Just Peace

Other Related Work

Conflict And Violence page

Democratic Peace page


Chapter 6

Same And Other;
Similarity And Difference*

By R.J. Rummel

And so iron will be mingled with silver, and brass with gold, and hence there will arise dissimilarity and inequality and irregularity, which always and in all places are causes of hatred and war.
---- Plato, The Republic


Consider the following sociological belief. Likes attract; birds of a feather flock together; or, in the words of the ancient Greek poets,1

God is ever drawing like towards like,
and making them acquainted.

The belief that similars attract, and its corollary, that opposites repel, has constituted a basic sociological explanation for love, friendship, solidarity, and affiliation, while the interaction of opposites has provided the basis for antagonism, hostility, strife, and conflict. Moreover, this belief became in early philosophy a general metaphor for explaining such puzzles as action at a distance (as of the moon on the tides) and magnetism, and in the hands of such ancient philosophers as Heraclitus became a cosmology. Indeed, down to our day, we continue to apply the basic concepts of attraction and repulsion to physical forces, and opposites or contradictions still play a strong role in some contemporary philosophies, as well as in the conflict helix.

However the simple notion that likes attract has not been universally accepted. In Lysis (213-223), Plato investigated whether friends were alike, or similarly good or congenial, and through the words of Socrates concluded "If neither the beloved, nor the loved, nor the like, nor the unlike, nor the good, nor the congenial, nor any other of whom we spoke--for there were such a number of them that I cannot remember all--if none of these are friends, I know not what remains to be said."

Hesiod, however, neither accepted that likes attract nor that the question was unresolved. Rather, he argued that likes repel. "Potter quarrels with potter, bard with bard, beggar with beggar" (Lysis, 215).

Although some have argued that unlikes attract and likes repel, sociological thought down to the present day has assumed that similarity is a cause of love, friendship, and solidarity, and the notion is incorporated in a variety of works. For example, James Davis (1966:82) posits "The more similar Person is to Other, the more Person will like Other." Or, stipulates Heider (1958:186) "p dissimilar to o induces p dislikes o; p tends to dislike a person different from himself." There is no need to multiply these examples, for the assumption can be found throughout contemporary literature.

And the assumption that likes attract has not gone without examination. Beginning with Plato, other philosophers such as Aristotle2 and Aquinas3 have subjected the question to analysis. And modern social scientists have empirically tested the notion that similarity causes or underlies love, friendship, or solidarity.4


The best contemporary sociological analysis of the belief that likes attract and consideration of the evidence pro and con is Sorokin's chapter "The Roles of Similarity and Dissimilarity in Social Solidarity and Antagonism," in his comprehensive Society, Culture, and Personality (1969). A summary of Sorokin's perspective and his conclusion will be helpful at this point, for his analysis is congenial to the understanding I am trying to develop.

Sorokin begins by pointing out three major theories on the question. One is that similarity between people leads to expressions of solidarity, such as marriage, or friendship; another, that dissimilarity leads to solidarity, as in the notion that unlikes attract; and finally, that both similarity and dissimilarity lead to solidarity.5 The latter includes theories of Durkheim, Tönnies, and Maclver, which divide solidarity into mechanical versus organic,6 gemeinschaft versus gesellschaft,7 community versus association.

Sorokin then considers the theory that similarity leads to solidarity in its most popular, contemporary manifestation: that like marries like. His survey of the empirical evidence tends to support this view. Generally, mates seem to be more alike than different.

But the evidence shows only a tendency, and as Sorokin notes, there is even reason to question this. First, the correlations between marriage and similarity are low. Second, the findings are not consistent regarding what constitutes similarity. The traits in one study related to marriage may be uncorrelated or negative in others. Third, in some studies marriage is related to dissimilarity, not similarity. Finally, these studies involve many arbitrary methodological assumptions and judgments. These considerations indicate to Sorokin that the empirical evidence hardly confirms the theory that similarity leads to solidarity.

Characteristically Sorokin, not content to rest on an empirical survey, uses it to introduce an analysis of the theory. He observes that those who have empirically investigated whether likes attract have not analyzed the meaning of "like." Perhaps similar people marry simply because they live close together; their choice is limited by physical distance. Since neighborhoods tend to be homogeneous in racial, religious, and socioeconomic characteristics, within this arena of choice marriages, by chance, would tend to be among likes. No mystical special attraction of likes need be postulated.

Moreover, propinquity is not the only factor in choice. Surely language, status, and race also limit or pattern inter action, thus serve as natural channels for one's choice of mate. Not similarity per se, but whom one meets makes for marriage.

Sorokin's point is a good one, but more recent studies have compensated for this to some extent. For example, a study of marriage patterns in Hawaii (Parkman and Sawyer, 1967) shows that those who are similar in racial-ethnic-cultural backgrounds tend to marry each other, in spite of a considerable mixing of racial-ethnic-cultural types (Japanese, Chinese, Korean, Okinawan, Filipino, Portuguese, Hawaiian, and so on) within the same neighborhoods, schools, and occupations, and similar language and status.

Regardless of these influences on choice, for Sorokin similarity itself seems to play little overall role in the happiness of a marriage. For some occupational groups, similarity seems important, whereas dissimilarity is significant for others. What may be most important for the happiness and the success of a marriage is the similarity in attitudes between the partners, but the evidence on this is inconsistent.

As final one-two punches, Sorokin argues that the universal taboos against marriages between close relations would not make sense if similarity were the crucial element in choice of mate and subsequent happiness. Moreover, marriages are usually between members of opposite sexes, important evidence against the theory that likes attract. Sexual differences, as any husband or wife knows, entail a wide range of physiological, emotional, and behavioral differences.8

Sorokin finally concludes that he must reject the theory that likes marry. And in consideration of the evidence and the foregoing points, he also rejects the opposite theory that dissimilarity leads to solidarity.

I must agree. I would soften the negative conclusion somewhat by saying that the sociocultural evidence and empirical studies do not substantiate nor invalidate the idea that similarity or dissimilarity underlies solidarity or antagonism. Our knowledge is inconclusive. I say this without accepting in total Sorokin's line of attack. For his analysis can be faulted for considering the meaning of similarity in a superficial way. But I return to this later.

Having cleared the terrain of false theories, Sorokin examines the theory that similarity and dissimilarity in combination underlie solidarity. Evidence for this is found in the observation that friends generally show a mixture of similar and dissimilar characteristics. The empirical tests of the relationship of similarity alone to friendship reveal that traits are mixed in their contribution. Moreover, marriage seems to require some dissimilarity (in sex and related characteristics) but some similarity in attitudes (toward children, handling money, etc.).

What is the nature of similarity and dissimilarity in general? After all, there are many different kinds, and all are not equally important or, in Sorokin's terms, virulent. No one characteristic by itself, whether age, race, height, weight, or education appears to generate solidarity or antagonism. For example, religion may be a central element in solidarity in some societies at some times, as in Europe in the Middle Ages; or it may be almost an irrelevant characteristic, as in contemporary Japan. What is virulent as a similarity or difference depends on the importance society attaches to that quality. A characteristic receives its virulence from the values attached to it by those involved.

Sorokin's point, central in his overall sociology, is that there are no objectively important things. Objective reality becomes socioculturally significant through our meanings, values, and norms. No objective dissimilarities or similarities can have significance, sui generis, in generating solidarity or antagonism. We must imbue characteristics with meaning and weight them with our values before they begin to take on importance to us. Thus there has been inconsistency in the empirical evidence about which characteristics (e.g., race or religion) are related to marriage or friendship.

Regarding the relation of similarity and dissimilarity to solidarity in general, Sorokin first says that the identity of the characteristics that comprise socially relevant similarity and dissimilarity depends on the sociocultural mentality, on their meanings and values. This constitutes the important similarity.

Second, if the characteristics are unimportant within a system of values, similarities or dissimilarities between them are unrelated to solidarity or antagonism. This particular conclusion goes far beyond Sorokin's analysis and evidence, and is questionable at its foundation. All factors that affect behavior, whether solidarity or antagonism, are not sociocultural. We have psychological needs, complexes, mechanisms that simply may be unrecognized by our sociocultural mentality. Aggressiveness as a complex of self-assertion and pugnacity needs and dominance and paranoidal temperaments, for example, may have no meaning or value within a culture, being an unconscious psychological disposition that neither the individual nor society recognizes. Yet the difference between two people in this disposition could well influence their interaction. If both are strongly aggressive, antagonism may result. Solidarity may require that one or both parties be characteristically nonaggressive. I am not saying that such is the case, only that Sorokin's characteristic reduction of reality to meanings and values is carried too far.

A third generalization is that an important sociocultural similarity generates solidarity if (a) the values for all are abundant and sufficient--they can be shared by all, such as God for Christians, pride for patriots, or truth for scholars; (b) the egos of the parties are fused into one "we," as in a family, intimate friendship, or close association (then the gain of one is the gain of the other, and even scarce values are shared); (c) and the relevant norms--standards or rules of conduct--are concordant.

Fourth, antagonism is generated by similarity if (a) values such as land, status, or food are scarce and the norms are permeated by an egoistic competition and discord; (b) and the norms are discordant and the values (e.g., becoming mayor of Honolulu) cannot be shared by all.

Fifth, if two people have quite dissimilar values, and those of one are considered unimportant by the other, neither solidarity nor antagonism is generated.

Sixth, similarity in values and norms, in the sense that they are considered equally important, generates antagonism if one affirms what the other denies. For example, a communist and an anticommunist may consider the same things to be important; but much of what one cherishes, the other may detest.

A final generalization is that solidarity is encouraged if the main values and norms are similar, and the secondary values are supplementary, or at least diverse or neutral. Thus a marriage between those of different races or religions can be solidary as long as the partners' main values are similar and their secondary values and norms are not discordant.

Sorokin then (1969:143) arrives at this basic conclusion about similarity and dissimilarity.

From the above it follows that the combination of a basic similarity in the main values and a concordance in the norms of the parties concerned, with a supplementary diversity in their secondary values, is the most conducive to solidary relationships, provided the main values are abundant, or are distributed by all the parties involved according to their concordant norms.

An opposite and diverse character of the values and norms of the parties, when they are considered important (positively or negatively), and in which the parties have no common system of values and norms, is the most conducive to the generation of intense antagonisms.

Thus Sorokin's analysis: it is sociologically insightful, comprehensive, and properly emphasizes the role of values and norms. Yet is not complete. A more careful consideration of the nature of similarity and difference is needed. The analysis should include another perspective on these two qualities--that of social distance. And consideration should be given to both the overall sociocultural context--the field--within which similarities and differences take on value and meaning, and a more specific understanding of values and norms.9


There is always some point of view in which everything is like every other thing....
---- Plato, Protagoras, 331

To say that likes attract, that solidarity is based on similarity, is to imply some meaning of "alike" or "similar." But what is this meaning? To say two people, or things for that matter, are the same, is to say what?

In answering, let me first make some distinctions. Consider that two people can be compared with respect to their qualities or quantities. Now "qualities" refer to the attributes, characteristics, and properties of people. They are the nonquantitative features by which we classify, categorize, and stereotype the profusion of manifestations that are another's field of expression. These qualities are nothing more than the latents underlying the transitory specifics, determinates, and manifestations that I discussed in Chapter 8 of The Dynamic Psychological Field, such as physical features, mannerisms, habits, and beliefs.

Two people may have similar, but not exactly the same, qualities, such as red hair. Like similar triangles, two people can share the same form but be quantitatively unequal. A baby, a short person, a tall person all can be fat, yet quantitatively they differ considerably. Therefore, of qualities we say that people are similar or alike, but not equal or unequal. Thus we categorize people as similarly whites, in spite of the wide quantitative differences (if we were to try to determine their average color with a spectroscope) among them.

Do qualities admit of degree, or can we say only that people are similar and dissimilar? Here there are two kinds of quality: binary and ordinal. Binary qualities are those that one either has or has not. Being alive, being a member of the Democratic Party, being a banker, being a father are all qualities that are either present or absent. There are then only two possibilities in the similarity or dissimilarity of people in their binary qualities, and if dissimilar, there is no question of degree. For example, if there are three people, i, j, and k, and only i is a governor, then i and j, and i and k are dissimilar in the quality of holding the office of governor. We cannot compare these two pairs to see whether one pair has more or less of this quality. Moreover, j and k are similar in that neither is a governor.

However some qualities allow more comparison than just the binary type. People may be compared on the basis of their power, prestige, responsibility, happiness, friendship, dominance, hunger, or beauty. For such qualities, we may be able to order people with regard to each other, so that this fellow is first, that fellow is second, and so on. Of course, we intuitively rank others on many qualities, such as beauty, power, and trustworthiness. And on such qualities when we say two people are similar, we mean that they share the same ranking. But unlike the binary qualities, when we say two people are dissimilar we do admit of some comparisons in dissimilarity. That is, in ranking on power, two people may be dissimilar but still rank close to each other, such as a mayor and a governor, a president and a vice president. Or they may be dissimilar in power and ranked at opposite extremes, like master and slave, general and private, king and serf. The qualities on which we can base comparisons of the dissimilarities of people I call ordinal.

Aside from dissimilarity in binary and ordinal qualities, people also can differ in quantity, in the amount of some measure that may be taken of them, such as height, weight, age, I.Q., income, years of education, size of home, reading speed, or size of family. On quantities, we say that people are equal or unequal.

Inequality clearly implies a degree or specification of an amount on which people can be compared in their inequality (the difference in age between Mary and Jim is ten years). There is a measurable quantitative difference that can be said to be twice as much or three times more than others. This level of specificity is sufficient for our purposes here. I could define quantitative differences entailing merely interval differences as distinct from those comprising a natural underlying zero point, but to no purpose.10

We have then qualities and quantities on which people may be respectively similar or dissimilar, equal or unequal. And dissimilarities can be binary or ordinal, and inequalities may differ by some amount or multiple. But the question still involves cutting across these distinctions. On what should two people be compared? On what qualities or quantities should we say they are similar or dissimilar, equal or unequal?

Here we get into the philosophical distinction between essence and accidents. The essence of a person is that which defines what he is. It is that without which he would not be a person, or Mary, or Catholic, or American. For example, the essence of a human being may be having reason or morality; the essence of an American may be independence, self-reliance, and a sense of equality. Clearly essence means a thing's nature. Accidents are then the qualities or quantities that are irrelevant to a thing's essence. So to human beings qua human beings, fingernail length, hair color, weight, power are accidents. They do not define a human's essence.

Clearly, then, if we are to compare people in order to say they are similar or dissimilar, equal or unequal, we want to compare their essences. We want to compare not the changing, ephemeral manifestations that from one moment are part of a person, but the invariant underlying latents that constitute you or me. But this brings us no closer to an answer to our question. For I have replaced "What shall we compare?" with "How do we determine essence?"11 An answer in part goes back to ancient philosophical concern with the nature of genus and species12 and in part to the more modern twist given this concern by William James in his treatment of discrimination and comparison in Chapter 13 of The Principles of Psychology (1890).

First, the essence of something constitutes that which all such things have in common. Whatever is the essence of a person, of social interaction, of conflict, is shared by all people, all social interaction, and all conflict. It was this essence, for example, of all social interaction that was my concern in Chapter 9 of Vol. 2: The Conflict Helix, where I argued that all social interactions share an orientation toward others.

Second, what is accident is something unique, like a person's biography, a car's rust spots, or a painting's detail. We have here the now familiar distinction between manifest and latent, but I want to pursue this distinction in a slightly different language more appropriate to our immediate concern.

We find that people are similar on one binary quality, dissimilar on another one, similar on this ordinal quality, dissimilar on that, and dissimilar in varying amounts on certain quantities. Indeed, in looking over the variety of qualities and quantities comprising us all, we may note that they can be divided into common and unique parts. Common is that which is shared by all, although we may be dissimilar or unequal in quality or amount. For example, size is common to all, but height and weight vary; power is common to all, as is a status. Moreover, there are also unique qualities and quantities, such as a person's smile, his voice, his likes and dislikes, his loves and hates, and his body measurements. By heredity, environment, experience, and individual will, each of us is a unique event in nature; but by heredity, environment, learning, society, and culture, each person is an example of human beings, of a culture, of a society, of an occupation, of a family. We are thus a composite: in confirming our commonness, in reflecting our individuality.

But as social scientists trying to understand social interaction and conflict, our concern is not with what is manifested only in one person, with the unrecurring, unpatterned, unique person, but with the common parts of his qualities and quantities, with people in general. At this point there are two ways to arrive at a framework for defining differences and similarities, and both end at the same place. In one we consider the common aspects of qualities and quantities, their variation, their patterns, and their constituent latent functions. This is the approach I took in The Dynamic Psychological Field, especially Chapters 10, 13, and 33, and I have taken it here, as well. But now my explicit concern is with the differences between two people. So let me choose the other route, which focuses on these distances.

We must find a way of discriminating between essence and accident, between our common and unique aspects. First, regardless of whether quality or quantity is essence or accident, let us compare the differences among all people on this dimension. I use the term comparison as a free variable ranging across binary and ordinal similarity and dissimilarity, and quantitative equality and inequalities. Thus two people will have an infinite variety of comparisons across all the qualities and quantities that could possibly characterize them.

Second, let us consider how these comparisons themselves vary across people. We will find that some comparisons are highly correlated, forming a pattern, a cluster. For example, there will be a pattern of differences associated with income, since people who differ in income usually differ in a variety of other characteristics, such as neighborhood, home, education, dress, and occupation. Similar examples could be given for those sharing the same religion, race, family, and so on. Thus the variation of people across a multitude of comparisons falls into patterns--intercorrelated clusters of comparisons.

Third, these patterns of comparisons define our common aspects or parts. And as we consider the variation among all comparisons involving all qualities and quantities, whether essence or accident, we will find that those involving our unique aspects remain unique.

We have in this a general epistemological route to defining commonness, the same route that Hume took in defining inductively cause and effect and natural law. Commonness is given us by the covariance of the comparisons between us. Those comparisons based on common aspects in our qualities or quantities will covary into empirical patterns. It is thus that we discern man and woman, father and child, rich and poor, powerful and weak. And it is thus that we can compare societies, as was done in Vol. 2: The Conflict Helix (Part IX).

But these relationships between comparisons, these covariances, are actually latent functions (Section 10.2 of Chapter 10 in Vol. 2: The Conflict Helix). They are the invariant common latents underlying all the manifest variation in the comparisons. And these common latent functions must be the components of a space of all the common aspects of qualities and quantities involved in these comparisons.

Before focusing on this space, let me review the logic. People have qualities and quantities that are partially shared with others and partially unique. We can hypothesize comparing people--considering their similarities and dissimilarities, equalities and inequalities--on all conceivable qualities and quantities, and then determining the covariance among all these comparisons. The common aspects of the comparisons will form patterns of regularity, while the unique comparisons will remain unrelated to any other. Now, these patterns, being common latent functions, thus define the components of a common space of comparisons. It is a space comprising not only quantitative comparisons, but binary and ordinal qualitative comparisons as well.13

But what have I bought? Of what use is this space? Remember that my ultimate concern is to have a basis for saying that two people are alike or different; and related to this, to determine what is essence and what is accident. I have shown that covariance is a route to defining essence; that the covariances among the multifold variety of comparisons between people are reducible to a common space. This space has the following value.

People are located in this space on the basis of their many common similarities and dissimilarities and equalities and inequalities. That is, their locations represent their mutual likeness on their essential characteristics, attributes, and properties. We have not only a solution to defining essence, but also to determining how and in what way people are alike or different.

The proximity of people in the space then defines their likeness. And since this is a metric space, we can use now a more precise measure of likeness, or of similarity-dissimilarity, equality-inequality--namely, the distance vector. This vector between any two people being compared represents the overall difference between them; and the orientation of this vector, its direction, defines the components of this difference.

Figure 6.1

For example, consider the hypothetical comparison space of Figure 6.1, spanned by two components, wealth and power. These constitute two patterns of covariance (Chapter 17 of Vol. 2: The Conflict Helix) underlying the diverse status comparisons among the hypothetical people shown (hollow dots). Within this space these people are distributed according to their wealth and power, and how close they are in this space defines their similarity on these two statuses. Two such persons, i and j, appear in the first quadrant along with the distance vector from j to i. Note the some people cluster together. These indicate or define those species or groups of people fairly close to each other on these comparisons relative to others: these clusters define types.

Henceforth, I use the term "distance" for the common similarity and dissimilarity, equality and inequality between two people in their essential characteristics, properties, or attributes. Distance stands for the magnitude of the distance vector in the comparison space just defined.

How do we know two people are alike? We assess the distance vectors between them.14 How do we know in which way they are alike? We assess the components of their distance vectors. How do we know how people group in their distances from each other? We determine empirically what clusters exist in this space, as was done for state-societies in Vol. 2: The Conflict Helix (Section 34.3 of Chapter 34). But here my interest is not empirical; rather, I want to assess theoretically likeness in its relation to solidarity or antagonism. So far, I have pushed the question up through layers of distinctions into common space. What remains to be assessed is the nature of the components of this space.

The common space, of course, is precisely the same as the sociocultural space (Chapter 13 of Vol. 2: The Conflict Helix) and the conflict space (Section 28.1 of Chapter 28 in Vol. 2: The Conflict Helix) latent in the conflict helix. For whether we begin with people as a field of expression defined by underlying latent functions as in Vol. 2: The Conflict Helix, or with comparisons between characteristics, properties, and attributes, as here, we finish with the same common space. These are two routes to the same point of view. Thus the distance vectors between people in their sociocultural space are the same distance vectors entailed in comparing people.15

Since the common space is our sociocultural space, the components are the same. These are the cultural components of language, religion-philosophy, ethics-law, science, and fine arts, and the social components of wealth, power, prestige, and class. The distances between people in the common space of all these components describe their likeness; the distances on these separate components, such as science or wealth, describe in which way two people are similar or different.

We now can deal more precisely with Sorokin's analysis of solidarity in the context of the conflict helix. First, however, I want to underline the ontology adopted here: what is similar or different is a point of view, a perspective on reality. Some emphasize the similarity of things, the genus, and strive toward defining the highest unity; some emphasize differences and analyze things into species and subspecies. Some see certain similarities as important, but others focus on different ones. All these perspectives are different actualizations of the potentials and dispositions existing in reality. In the light of our interests, of our problems, of our intentions, we confront the reality of other people with a framework of unity or differences; other people confront us with their power to manifest a specific reality. The balance between our perspective on unity and differences and the power of this reality constitute our perception.

For example, the particular component distances between people in sociocultural space will shift as we rotate its components. Yet nothing in the empirical comparisons between people will uniquely specify a rotation for this space. Consequently, an infinitude of perspectives on likeness all can be compatible with the same manifest comparisons. Of course the space is invariant and the distance across the whole space is invariant of rotations: only one space and overall distance are consistent with the same data. But even on this we get an invariant space and overall distance only if we assume the space to be Euclidean.

Given, then, that our view of likeness or difference is partially a perspective with which we confront nature, a perspective that is a condition of nature's intelligibility, how do we as social scientists investigating social interaction select an appropriate perspective? We do so in terms of a number of criteria, such as the moral implications, the comprehensiveness, the power of a perspective to make social reality intelligible, and simplicity. Above all, however, we select that perspective which helps us resolve our problems and answer our questions. In terms of my interest here, we select that perspective which best explains social interaction, such as solidarity and antagonism. The justification of the perspective I am adopting does not lie in the perspective (the space, distances, components) itself, therefore, but in its utility in helping us to understand interaction.


In previous sections I presented Sorokin's analysis of the relationship of similarity and difference to solidarity and antagonism. His solution was to attempt throughout to subjectify differences, to treat them as a matter of individual meanings, values, and norms. With this conclusion in mind, I tried to provide a useful perspective for understanding similarity and difference, especially in relation to norms and values. The final questions now concern the precise meaning of norms and values in the light of our common sociocultural space, and the relationship of the whole question of individual similarities and differences to the conflict helix.

What is a norm? In general it is a standard or role that guides or regulates behavior. Is there some other meaning, however, that is intrinsic to sociological analysis? When we consider what Sorokin (1969, Chapter 42) means, we find that norms define our highest moral standards, such as the Ten Commandments, or our most practical rule, such as those of etiquette, methodology, or good gardening (p. 47). This statement is clearly too broad for our purposes, since to say that interaction is dependent on similarities and differences in norms hardly focuses our comprehension well. We are then forced to ask about the kind of norms involved.

Perhaps others may provide a more useful meaning to norms. The Dictionary of Sociology and Related Sciences (1970) defines a norm as any "socially sanctioned mode or condition of behavior." This definition is even less helpful than Sorokin's. Three other definitions give some assistance. For Merton, norms are regulations about what are acceptable ways of achieving goals.16 For Peter Blau (1967, especially p. 255), norms are standards of permissible conduct; and for George Homans (1961:46), a norm is "a statement made by a number of members of a group, not necessarily all of them, that the members ought to behave in a certain way in certain circumstances." Here we have norms as behavioral oughts, as regulators of interaction. They are our social ethics. I believe that norms as understood in this way fit precisely with Sorokin's conclusion about the relation of norms to solidarity and antagonism.

But, is not a norm also our expectation? Do we not expect others to behave according to our norms? In the previous volume (Section 14.3 of Chapter 14 in Vol. 2: The Conflict Helix) I defined our expectations as the anticipated outcome of our behavior. We behave toward another person in the expectation that certain behavior or responses will result. This expectation has two aspects. First, there is an empirical ingredient, a learned expectation of another gained from prior interaction with him. A women may have learned, for example, that her husband becomes irritated if she burps at mealtime, her boss at work responds well to compliments, or her son balks at demands.

Clearly this aspect of expectations may not be normative. No ought or standard may be involved in our expectations as a learned outcome. The other may consistently act contrary to our norms or those of society, and with resignation, or with simply a wish to get along, we come to expect and accept this.

However there is one more ingredient, namely, our culturally learned expectations of another's behavior. We expect another to follow certain etiquette in responding to us, to obey certain implicit rules of the game, to act according to a certain morality. Society and culture are crosscut by these norms, which regulate behavior, and our expectations manifest them. For based on minimal information about another, we initially expect him to act as norms dictate. Only as we gain experience, do our expectations become more personalized and encompass deviation from cultural norms.

We can define two kinds of expectation, therefore, which combine to form our expected outcomes. Our social expectations are the outcome we expect according to our sociocultural and group norms. The norms serve as our generalized prediction of the responses of others; they provide us with an ability to swim through a sea of strangers. Our personal expectations comprise what experience leads us to anticipate about the response to our behavior from a particular other (such as a specific student named John) or a similar class of others (all students).

Social norms and social expectations thus are linked. Norms define the proper behavior in a society; our belief that people will so behave are our social expectations. Then the similarities or differences between two people in their social expectations of each other really reflect, to use Sorokin's terms, their concordance in norms.

And consider. Is not the structure of expectations and its congruence with an underlying balance of powers the measure of distance in expectations between individuals? Clearly it is. For it is through conflict that different expectations are corrected. And it is the structure of expectations that forms a congruence of expectations sufficient for cooperative-solidary-interactions.


Mythology, religion and philosophy ... are alike in that each has as its function the interpretation of experience in terms which have human value.
---- Joseph Wood Krutch, The Modern Temper, Vol. I

Those studying social interaction, and especially conflict, have long recognized the importance of similarities or differences in values to harmonious or hostile relations. But here also we must ask, What are values? Sorokin is of no help, because he treats values as an implicit aspect of meanings and norms. Nor are other sociologists, for there is considerable confusion about values across the literature. Clearly a basic philosophical analysis is required. This may help in our understanding of values and in developing a conceptual framework consistent with our perspective on social interaction and the conflict helix.

When we say that Bob is good, the food is tasty, the government is bad, the kitten is cute, and the idea stinks, we are expressing values. The things named are being weighed by us; as concepts they are evaluated. Adjectives like good, bad, and ugly express our feelings, our positive or negative sentiments toward things. To say something is bad (or terrible or horrible or outrageous) is to express negative affect, an emotional repulsion; to assert goodness of some sort, however, is to indicate that an attraction exists. We need not rely on intuition alone, however.

On the empirical level, analyses of concepts have shown consistently and cross-culturally that such an evaluation dimension is one of the kinds of conceptual meaning with which we endow things.17

One meaning of value is our emotive feeling toward things. Values, in part, refer to our disposition to endow that which we perceive with meanings like ugly or beautiful. But are not all values emotive? When we say that it is wrong to kill or lie and right to help others and turn the other cheek, are we not saying that these behaviors are good or bad, that they attract or repulse us?

Such is the ethical theory of many positivists, like A. J. Ayer (1946), who believe that all morality is reducible to emotive statements. I will not become involved in this popular argument here, reserving that fun for Vol. 5: The Just Peace. However I can at least establish a line of attack incorporating another meaning of values--that involving ethics--which should enable me to clarify the relationship of values to interaction.

Although the matter is not undisputed,18 the philosophical consensus is that Hume guillotined propositions about reality into two types:

In every system of morality which I have hitherto met with, I have always remark'd that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surpriz'd to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is, however, of the last consequence. For as this ought or ought not, expresses some new relation or affirmation, 'tis necessary that it should be observ'd and explain'd; and at the same time that a reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it.19

That is, statements about reality are of two logically distinct kinds. Those that describe or state what is, and those that state what ought to be. One cannot logically move from statements about what is to statements about what should be.

Statements about what should or ought to be comprise, however, our ethics, our morality. People should not kill; husbands should be faithful; children should be educated. But by Hume's analysis and contemporary philosophical opinion, such statements are logically distinguishable from beliefs about reality. Similarly, in Kant's analysis, reality as phenomena are what we experience--the realm of cause and effect, of natural laws. The realm of morality, of ethics, is nonempirical. It is the arena of pure reason, the domain of practical judgment.

We can therefore define a realm of values--our oughts, our morals, our ethics. Values of these kinds are of three types. First, our sociocultural norms or standards are the oughts that govern and regulate our social interaction, as already discussed. Second, instrumental values are the oughts that define the means, ways, routes, methods, structures, and so on for achieving some fundamental value. Thus democracy may be an instrumental value for achieving freedom, or the Good; welfare, for alleviating hardship; war, for protecting national values. Finally, the fundamental values, the ends, are usually nonnegotiable and categorical. These are the basic oughts for which one is willing to sacrifice even himself. God, truth, freedom, equality, humanity, law, family, country have historically been the foci of such imperatives. They are our most basic ethics.

If values are norms, instrumental, or fundamental, how do we know whether people are alike or different in their values? To begin with, such values are subjective. Each of us will rate them in our own way, as a part of our psychological field.

Sociocultural norms can be identified with expectations. Is there such a one-to-one translation for values? First, I must determine what aspects, structures, or processes of the psychological field correspond to these values. Obviously important is the cultural matrix, which endows stimuli with meaning, values, and norms in the act of perception; it gives the initial weighting to reality. We do not see just a complex of lines, colors, and shapes, nor just a painting. We see a beautiful or ugly painting. We apprehend the world evaluatively. The cultural matrix defines our emotive values.

This creates a problem. For "differences or similarities between people in their cultural matrices" carries no empirical meaning. Certainly the cultural matrix underlies our perspectives, but "a perspectival distance" is too broad a term, since perspectives include personality and station, as well as the cultural matrix.

There is a solution, however, as indicated in the previous sections. If any psychological structure is suffused with emotive values, it is the attitudinal lattice that connects our basic needs with our day-to-day dispositions. Reflecting our emotive values are our attitudes toward abortion or a politician; our desire for a particular car or a higher salary; and our goals of helping the disadvantaged or learning to play the piano. Indeed, our positive and negative interests, which are simply our attitudes and their power to become manifest, are saturated with evaluation. To want very much to eat a steak, to buy a painting by Degas, to have the public water supply fluoridated, and to avoid inflation is to manifest attraction or repulsion with respect to such goals. And what else are emotive values?

For these values, then, the means of considering similarity and differences is clear: it is the distance vector between individuals in their cultural space. The components of this vector reflect fundamental differences in religion, science, ethics, and so on. But attitudinal differences also lie along this vector, for the larger the distance between individuals in cultural space, the more likely that their emotive attitudes are different, and the closer in this common space, the more likely they are to share the same attitudes.

Interests are activated attitudes. Therefore if the attitudinal differences and similarities lie along the distance vectors, this will also be true of interests. Distances in cultural space reflect distances in interests. And since interests are the basic motivational unit, the visible surface of our needs, sentiments, and roles, distances in interests also measure distances in needs like security, sentiments like self-esteem, and role attitudes.

This leaves us still with instrumental and fundamental values, with comparing the value hierarchies of individuals. The fundamental values are a person's ultimate regulating principles. Are these also reflected in interests? Of course. The ends and means of our interests are surely regulated by moral judgment, and their content manifests our beliefs about killing, stealing, honor, loyalty, freedom, equality, and so on for however the Good is defined.

Here, however, we can be more specific about the nature of these interests than in the case of emotive values. Distinct interests (i.e., attitudes and their power to manifest behavior) coalesce as they share a common orientation toward humanity, society, and nature. Our interests combine as they reflect, say, a faith in the scientific method and empirical truth, a belief in the Buddhist creed, an appreciation and love of Japanese landscape painting. There is one pattern, however, one interrelated configuration of interests that consists entirely of values. These are attitudes toward the law's content and basic questions of moral personal and public policies. Should prostitution or the sale of pornographic materials be legal? Should there be capital punishment? Should there be amnesty for war deserters? Such are the stuff of politics and law, but our views on them, our attitudes that "I want the President's authority to make war taken away" or "I want to fight for Israel against the Arabs," are dispositions directly growing out of our fundamental values.

One component of intentional space involves these interests: the ethics-law component of our cultural subspace. Our location on this component--whether our ethics-law is hedonistic, transitory, relativistic, and totally sensory or a rational set of inflexible, God-given ideals--defines our fundamental values. Therefore to compare two people in their basic values, to assess their similarity or dissimilarity in ultimate oughts, is to consider their distance on the ethics-law component of cultural space.

Finally, there are the instrumental values that guide us in the achievement or maintenance of our fundamental values. Clearly, instrumental values as technical norms regulating the safe operation of a car (we ought to fasten our seat belt), doing scientific research (we ought to follow the scientific method to discover truth), or administering welfare are reflected in our varied interests. For all associated attitudes have an in-order-to aspect. I want to watch TV in order to relax or laugh or be entertained; I want to have public rapid transit in order to alleviate traffic congestion; I want to see an atmosphere of détente with the Soviet Union in order to lessen the risk of nuclear war.

But most attitudes involve instrumental values that as technical norms are relatively unimportant in comparing people, except perhaps methodologists (who may become quite emotional over whether one should use, say, factor analysis, causal analysis, or multidimensional scaling). There are, however, instrumental values that are fundamental values in their own right as well. Although socialism may be considered to be a means to greater human happiness, it itself may become a pseudo-ultimate value; although democracy may be considered to be a means to preserving liberty, it can become another pseudo-ultimate. And so on for such instrumental values as public education, government regulation of business, or a minimum wage; churches, universities, or courts; communism, liberalism, or anarchism.

From this it should be clear that like our ultimates, instrumental values fundamentally mirror our statuses and our class membership--our social relationships. Our fundamental values are cultural in origin, but our means to satisfy them become our society, our division of labor. And the major cleavages in this society, which are between rich and poor, powerful and weak, honored and ignored, and especially, rulers and ruled, determine rewards. Most within the same culture share a desire for the same kind of rewards and benefits, but who gets how much, and when, are matters of class, of authoritative instrumentalities. Therefore instrumental attitudes are reflected in the distance vectors between individuals in their common social space. And again, people's interests lie along these distances.

The political formulas that move people are basically an amalgam of fundamental and instrumental values, of culture and society, of ethics, of status. And of class. Politics and political issues lie at the center of societies and their cultures. These are questions of power, authority and class, wealth and prestige. Who will lead? How will the leader or leaders be selected? What laws should be passed? How will privileges and sanctions be distributed? More concisely, who gets what, how, and when? Whether we are talking about the politics of universities, businesses, unions, cities, or states, or international relations, as they engage our interests, they fundamentally engage our basic ethics. Our ethics--our fundamental oughts--endow our political interests with direction and strength. To know that a person is a communist, a libertarian, a Nazi, a Roosevelt liberal, a Maoist is to have bracketed his fundamental ethical-legal values.20 Political distances reflect such underlying similarities and differences as political types correspond to different societies and cultures (Section 34.2 of Chapter 34 in Vol. 2: The Conflict Helix).


We can now return to Sorokin's analysis of similarity and dissimilarity and his fundamental conclusion that the antagonism between individuals depends on the basic similarity in their main values and the concordance in their norms. First, Sorokin's assumption that the important differences and similarities are in meanings, values, and norms, and not in objective characteristics, is an assumption of the conflict helix. For the space from which conflict stems is the common sociocultural space of meanings, values, norms, statuses, and class. The differences significant to social conflict define social relationships. And such are determined by the relative locations of individuals in their common sociocultural space, that is, their conflict space.

Second, the significant similarities and differences are measured by the distance vectors between individuals on the common components of the sociocultural space. These are the distances in religion, in science, in ethics, and so on for the cultural components, and in statuses and class for the social components. These distance vectors in sociocultural space define likes and unlikes, same and other, similarity and dissimilarity.

Third, attitudes, thus interests, lie along the distance vectors between individuals in their common sociocultural space. The different locations of individuals in this space reflect different interests; and the further apart the more likely it is that their interests will be opposed. In terms of the conflict helix, distances in sociocultural space and mutual awareness of them define the structure of conflict. The transformation of attitudes into interests creates the situation of conflict.

Fourth, the concordance of norms between individuals in interaction is a function of their structure of expectations. It is a function of the congruence between this structure and the underlying balance of capabilities, interests, and wills.

Fifth, according to Sorokin, antagonism is generated by the 46 opposite and diverse character" of the important values and norms and "in which the parties have no common system of values and norms." From the previous analysis, we can now rephrase this as follows. Antagonism will be generated between people as they are distant in their expectations and interests. Expectations define general norms, and interests cover the emotive, instrumental, and fundamental values. Also, the degree to which two people have a common system of values and norms is actualized in their interests, their balance of powers, and their structure of expectations. With no common interests and social expectations, there can be no common system of values and norms.

Finally, the antagonism between individuals as a function of their expectations and interests, similarities and differences, depends on the conflict process. It depends on their mutual location in sociocultural space, on their awareness of the distances involved, on the activation of their interests, on their capabilities and wills, and on their balancing (which establishes concordant norms, a congruent structure of expectations).

In sum, Sorokin's conclusions are embodied in the conflict helix. The virtue of the helix is that it provides a clear definition of similarity and difference, taking into account qualitative and quantitative differences and imbedding these differences in an overall process of conflict that includes also culture, status, class, needs, attitudes, dispositions, interests, perception, power, and expectations. Thus the functioning of similarities and differences between individuals is understood as part of a sociocultural field. Differences alone do not create antagonism. But they are necessary ingredients in the antagonistic process, in the conflict helix. 


* Scanned from Chapter 6 in R.J. Rummel, Conflict In Perspective, 1977. For full reference to the book and the list of its contents in hypertext, click book. Typographical errors have been corrected, clarifications added, and style updated.

1. Plato, Lysis, 214.

2. Nicomachaen Ethics, Book Vill.

3. Summa Theologica, Part I of Second Part, O. 27, Art. 3.

4. See, for example, the empirical references mentioned in Davis (1966), Heider (1958), Burgess and Wallin (1943), Sorokin (1969, Chapter 7), and Berelson and Steiner (1964:305-309, 313).

5. Sorokin's categorization might resemble the stock market analyst's prediction, "The market will go up or down, or it could remain the same." Rather than developing a logical classification, Sorokin is inductively organizing the thinking of sociologists. To be logically complete, one also should pose the theory that both similarity and dissimilarity are independent of solidarity (or antagonism). No major thinker or scientist has proposed this, however; although by implication Plato came close to it through Socrates' inability to uncover the basis of friendship.

6. Mechanical solidarity is that of primitive or traditional societies based on similarity; organic is that of modern industrial societies characterized by extensive division of labor, which are sewn together by these differences. See The Division of Labor in Society (Durkheim, 1933).

7. Tönnies lumps all kinds of different elements into these two distinctions, which makes it difficult to distinguish them. In essence, however, it seems that gemeinschaft is a pattern of interaction involving fellowship, community, and love, whereas gesellschaft includes contractual and utilitarian types of interaction. Traditional sacred societies would be characterized by gemeinschaft; industrial secular societies by gesellschaft. See his Gemeinschaft and Gesellschaft.

8. For example, although men and women may use the same words, their modes of expressing these words through sentence structure, emphasis, and gesticulation are quite distinct.

9. In spite of their central role, values and norms as concepts are not subjected to any thoroughgoing analysis by Sorokin. He treats them as primitive terms. At the general macrolevel comfortable to Sorokin, this is consistent with his conclusions. The problem comes in subjecting a generalization involving values to a microanalysis, for then the apparently clear meaning of the generalization evaporates. This is because an analytic treatment of values or ethics would reveal a profusion of independent distinctions. For example, value can refer to our disposition or sentiment toward something (whether we like it), an instrumental norm (we value money for what it can buy), or a basic end (the Good).

10. This would distinguish the so-called interval and ratio scales of measurement.

11. It is a common contemporary position to argue that things have no essence, but only relationships. Therefore, what a thing is depends on the relationships of concern to us. I disagree. To say that a thing is defined by its relationships is to imply some thing that exists apart from the relationship. My argument on this is in Chapter 34 of The Dynamic Psychological Field.

12. In the context of a concern with same and other, similarity and difference, see, for example, Aristotle, Categories, 19-35.

13. Some mathematically oriented readers may question this. But we can denote binary qualities as zero or unity, and then the comparison between two people on this yields a difference of either zero (if they are similar) or plus or minus unity (if they are dissimilar); we can scale ordinal qualities and then compare two people by taking the ordinal difference. Consider next a matrix whose columns define the various qualitative-quantitative comparisons being made (as in religion, nationality, age, income, and education) and whose rows refer to specific pairs, such as John and Mary, Bill and Jean. Finally, consider a symmetric matrix comprising all the covariances (correlations) among these qualitative and quantitative comparisons. Then the eigenvectors of this matrix will approximate the latent functions of which I speak, and the eigenvectors in their totality will span the common space. Mathematically this will be a Euclidean space defining qualitative and quantitative comparisons.

14. The logic used here, the defining of a common space and distances between people in this space, underlies also a variety of approaches to similarities and differences being taken across the sciences. For example, pattern-search techniques being developed in mathematics, communications, and physics use similar approaches. See, for example, Watanabe (1969). Moreover, quantitative taxonomists in biology, trying to overcome the difficulties of the intuitive, genus-species approach, are employing the same approach to assess likeness that I am using. Foremost among them are Sokal and Sneath (1963), who recommend and elaborate methods identical to those I use in Dimensions of Nations (1972).

Multidimensional scaling and factor analysis are also similar logics for comparing individuals (Rummel, 1970, Chapter 22, on "Distance"; and "Understanding Factor Analysis").

15. Specifically, one can begin empirically with attributes of individuals and define mathematically a space and its dimensions, the location of individuals within it, then their mutual distances. Or, one can begin with their differences and similarities on the attributes and mathematically define the space containing them as distances. In both cases the resulting space and distances are similar (see Rummel, 1970, Chapter 22).

16. See Merton (1957), particularly page 133, in reference to institutionalized norms.

17. See Osgood et al. (1957) and Miron and Osgood (1966). Besides the evaluative, two major dimensions of conceptual meaning are potency and activity.

18. See, for example, Hudson (1969).

19. Treatise, L. A. Selby-Bigge's edition, p. 469.

20. Of course those sharing the same political ideology may differ with respect to their honesty and their willingness to use unethical and immoral means to achieve their political goals. I am assuming that such differences are reflected somewhat by political interests, but mainly by the nonpolitical interests representing the emotive and instrumental values of a person.

For citations see the Vol. 3: Conflict in Perspective REFERENCES

Go to top of document