The Functions of Stratification

KINGSLEY DAVIS AND WILBERT E. MOORE

Some Principles of Stratification
In a previous paper some concepts for handling
the phenomena of social inequality were pre-
sented.' In the present paper a further step in
stratification theory is undertaken-an attempt
to show the relationship between stratification
and the rest of the social order.' Starting from the
proposition that no society is "classless;' or un-
stratified, an effort is made to explain, in func-
tional terms, the universal necessity which calls
forth stratification in any social system. Next, an
attempt is made to explain the roughly uniform
distribution of prestige as between the major
types of positions in everysociety. Since, however,
there occur between one society and another
great differences in the degree and kind of stratifi-
cation, some attention is also given to the variet-
ies of social inequality and the variable factors
that give rise to them.
Clearly, the present task requires two different
lines of analysis-one to understand the univer-
sal, the other to understand the variable features
of stratification. Naturally each line of inquiry
aids the other and is indispensable, and in the
treatment that follows the two will be interwoven,
although, because of space limitations, the em-
phasis will be on the universals.
Throughout, it will be necessary to keep in
mind one thing-namely, that the discussion re-
lates to the systemof positions, not to the individ-
uals occupying those positions. It is one thing to
ask whydifferent positions carry different degrees
of prestige, and quite another to ask how certain
individuals get into those positions. Although, as
the argument will try to show, both questions are
related, it is essential to keep them separate in our
thinking. Most of the literature on stratification
has tried to answer the second question (particu-
larly with regard to the easeor difficultyof mobil-
ity between strata) without tackling the first. The
first question, however, is logically prior and, in
the case of any particular individual or group,
factually prior.
The Functional Necessity of Stratification
Curiously the main functional necessity explain-
ing the universal presence of stratification is pre-.
cisely the requirement faced by any society of
placing and motivating individuals in the social
structure. As a functioning mechanism a society
must somehow distribute its members in social
positions and induce them to perform the duties
of these positions. It must thus concern itself with
motivation at two different levels: to instill in the
proper individuals the desire to fill certain posi-
tions, and, once in these positions, the desire to
perform the duties attached to them. Eventhough
the social order may be relatively static in form,
there is a continuous process of metabolism as
new individuals are born into it, shift with age,
and die off. Their absorptioninto the positional
system must somehow be arranged and moti-
vated. This is true whether the systemis competi-
tive or non-competitive. A competitive system
gives greater importance to the motivation to
achieve positions, whereas a non-competitive sys-
39
40
tem gives perhaps greater importance to the mo-
tivation to perform the duties of the positions;
but in any systemboth types of motivation are re-
quired.
If the duties associated with the various posi-
tions were all equally pleasant to the human or-
ganism, all equally important to societal survival,
and all equally in need of the same ability or tal-
ent, it would make no difference who got into
which positions, and the problem of social place-
ment would be greatly reduced. But actually it
does make a great deal of difference who gets into
which positions, not only because some positions
are inherently more agreeable than others, but
alsobecause some require special talents or train-
ing and some are functionally more important
than others. Also, it is essential that the duties of
the positions be performed with the diligence that
their importance requires. Inevitably, then, a so-
ciety must have, first, some kind of rewards that it
can use as inducements, and, second, some wayof
distributing these rewards differentially according
to positions. The rewards and their distribution
become a part of the social order, and thus give
rise to stratification.
One may ask what kind of rewards a society has
at its disposal in distributing its personnel and se-
curing essential services. It has, first of all, the
things that contribute to sustenance and comfort.
It has, second, the things that contribute to hu-
mor and diversion. And it has, finally, the things
that contribute to self respect and ego expansion.
The last, because of the peculiarly social character
of the self, is largely a function of the opinion of
others, but it nonetheless ranks in importance
with the first two. In any social system all three
kinds of rewards must be dispensed differentially
according to positions.
In a sense the rewards are "built into" the posi-
tion. They consist in the "rights" associated with
the position, plus what may be called its accom-
paniments or perquisites. Often the rights, and
sometimes the accompaniments, are functionally
related to the duties of the position. (Rights as
viewed by the incumbent are usually duties as
viewed by other members of the community.)
However, there may be a host of subsidiary rights
and perquisites that are not essential to the func-
tion of the position and have only an indirect and
II I Forms and Sources 01 Slratllication
symbolic connection with its duties, but which
still maybe of considerable importance in induc-
ing people to seek the positions and fulfil the es-
sential duties.
If the rights and perquisites of different posi-
tions in a society must be unequal, then the soci-
ety must be stratified, because that is precisely
what stratification means. Social inequality is
thus an unconsciously evolved device by which
societies insure that the most important positions
are conscientiously filled by the most qualified
persons. Hence every society, no matter how sim-
ple or complex, must differentiate persons in
terms of both prestige and esteem, and must
therefore possess a certain amount of institution-
alized inequality.
It does not follow that the amount or type of
inequality need be the same in all societies. This is
largely a function of factors that will be discussed
presently.
The Two Determinants of Positional Rank
Granting the general function that inequality
subserves, one can specify the two factors that de-
termine the relative rank of different positions. In
general those positions convey the best reward,
and hence have the highest rank, which (a) have
the greatest importance for the society and (b) re-
quire the greatest training or talent. The first fac-
tor concerns function and is a matter of relative
significance; the second concerns means and is a
matter of scarcity.
Differential Functional Importance. Actually a
society does not need to reward positions in pro-
portion to their functional importance. It merely
needs to give sufficient reward to them to insure
that they will be filled competently. In other
words, it must see that less essential positions do
not compete successfully with more essential
ones. If a position is easily filled, it need not be
heavily rewarded, even though important. On the
other hand, if it is important but hard to fill, the
reward must be high enough to get it filled any-
way. Functional importance is therefore a neces-
sary but not a sufficient cause of high rank being
assigned to a position.'
TheFunctions of Stratification
Differential Scarcity of Personnel. Practically all
positions, no matter how acquired, require some
form of skill or capacity for performance. This is
implicit in the very notion of position, which
implies that the incumbent must, by virtue of his
incumbency, accomplish certain things.
There are, ultimately, only two ways in which a
person's qualifications come about: through in-
herent capacity or through training. Obviously,
in concrete activities both are always necessary,
but from a practical standpoint the scarcity may
lie primarily in one or the other, as well as in
both. Some positions require innate talents of
such high degree that the persons who fill them
are bound to be rare. In many cases, however, tal-
ent is fairly abundant in the population but the
training process is so long, costly, and elaborate
that relatively fewcan qualify. Modern medicine,
for example, is within the mental capacity of
most individuals, but a medical education is so
burdensome and expensive that virtually none
would undertake it if the position of the M.D.
did not carry a reward commensurate with the
sacrifice.
If the talents required for a position are abun-
dant and the training easy, the method of acquir-
ing the position may have little to do with its du-
ties. There may be, in fact, a virtually accidental
relationship. But if the skills required are scarce
by reason of the rarity of talent or the costliness of
training, the position, if functionally important,
must have an attractive power that will draw the
necessary skills in competition with other posi-
tions. This means, in effect, that the position
must be high in the social scale-must command
great prestige, high salary, ample leisure, and the
like.
How Variations Are to Be Understood. In so
far as there is a difference between one system of
stratification and another, it is attributable to
whatever factors affect the two determinants of
differential reward-namely, functional impor-
tance and scarcity of personnel. Positions impor-
tant in one society may not be important in an-
other, because the conditions faced by the
societies, or their degree of internal development,
may be different. The same conditions, in turn,
may affect the question of scarcity; for in some
41
societies the stage of development, or the external
situation, may wholly obviate the necessityof cer-
tain kinds of skill or talent. Any particular system
of stratification, then, can be understood as a
product of the special conditions affecting the
two aforementioned grounds of differential re-
ward.
Major Societal Functions
and Stratification
Religion
The reason why religion is necessary is apparently
to be found in the fact that human society
achieves its unity primarily through the posses-
sion by its members of certain ultimate values
and ends in common. Although these values and
ends are subjective, they influence behavior, and
their integration enables the society to operate as
a system. Derived neither from inherited nor
from external nature, they have evolved as a part
of culture by communication and moral pressure.
They must, however, appear to the members of
the society to have some reality, and it is the role
of religious belief and ritual to supply and rein-
force this appearance of reality. Through belief
and ritual the common ends and values are con-
nected with an imaginary world symbolized by
concrete sacred objects, which world in turn is re-
lated in a meaningful wayto the facts and trials of
the individual's life. Through the worship of the
sacred objects and the beings they symbolize, and
the acceptance of supernatural prescriptions that
are at the same time codes of behavior, a powerful
control over human conduct is exercised, guiding
it along lines sustaining the institutional struc-
ture and conforming to the ultimate ends and
values.
If this conception of the role of religion is true,
one can understand why in every known society
the religious activities tend to be under the charge
of particular persons, who tend thereby to enjoy
greater rewards than the ordinary societal mem-
ber. Certain of the rewards and special privileges
may attach to only the highest religious function-
42
aries, but others usually apply, if such exists, to
the entire sacerdotal class.
Moreover, there is a peculiar relation between
the duties of the religious official and the special
privileges he enjoys. If the supernatural. world
governs the destinies of men more ultimately
than does the real world, its earthly repre-
sentative, the person through whom one may
communicate with the supernatural, must be a
powerful individual. He is a keeper of sacred tra-
dition, a skilledperformer of the ritual, and an in-
terpreter of lore and myth. He is in such close
contact with the gods that he is viewed as possess-
ing some of their characteristics. He is, in short, a
bit sacred, and hence free from some of the more
vulgar necessitiesand controls.
It is no accident, therefore, that religious func-
tionaries have been associated with the very
highest positions of power, as in theocratic re-
gimes. Indeed, looking at it from this point of
view, one may wonder why it is that they do not
get entire control over their societies. The factors
that prevent this are worthy of note.
In the first place, the amount of technical com-
petence necessary for the performance of reli-
gious duties is small. Scientificor artistic capacity
is not required. Anyone can set himself up as en-
joying an intimate relation with deities, and no-
body can successfullydispute him. Therefore, the
factor of scarcity of personnel does not operate in
the technical sense.
One may assert, on the other hand, that reli-
gious ritual is often elaborate and religious lore
abstruse, and that priestly ministrations require
tact, if not intelligence. This is true, but the tech-
nical requirements of the profession are for the
most part adventitious, not related to the end in
the same way that science is related to air travel.
The priest can never be free from competition,
since the criteria of whether or not one has genu-
ine contact with the supernatural are never
strictly clear. It is this competition that debases
the priestly position below what might be ex-
pected at first glance. That is why priestly prestige
is highest in those societies where membership in
the profession is rigidly controlled by the priestly
guild itself. That is why, in part at least, elaborate
devices are utilized to stress the identification of
the person with his office-spectacular costume,
II I Forms and Sources of Slralilicalion
abnormal conduct, special diet, segregated resi-
dence, celibacy, conspicuous leisure, and the like.
In fact, the priest is alwaysin danger of becoming
somewhat discredited-as happens in a secular-
ized society-because in a world of stubborn fact,
ritual and sacred knowledge alone will not grow
crops or build houses. Furthermore, unless he is
protected by a professional guild, the priest's
identification with the supernatural tends to pre-
clude his acquisition of abundant worldly goods.
As between one society and another it seems
that the highest general position awarded the
priest occurs in the medieval type of social order.
Here there is enough economic production to af-
ford a surplus, which can be used to support a
numerous and highly organized priesthood; and
yet the populace is unlettered and therefore cred-
ulous to a high degree. Perhaps the most extreme
example is to be found in the Buddhism of Tibet,
but others are encountered in the Catholicism of
feudal Europe, the Inca regime of Peru, the Brah-
minism of India, and the Mayan priesthood of
Yucatan. On the other hand, if the society is so
crude as to have no surplus and little differentia-
tion, so that every priest must be also a cultivator
or hunter, the separation of the priestly status
from the others has hardly gone far enough for
priestly prestige to mean much. When the priest
actually has high prestige under these circum-
stances, it is because he also performs other im-
portant functions (usually political and medical).
In an extremely advanced society built on sci-
entific technology, the priesthood tends to lose
status, because sacred tradition and supernatural-
ism drop into the background. The ultimate
values and common ends of the society tend to be
expressed in less anthropomorphic ways, by offi-
cials who occupy fundamentally political, eco-
nomic, or educational rather than religious posi-
tions. Nevertheless, it is easily possible for
intellectuals to exaggerate the degree to which the
priesthood in a presumably secular milieu has
lost prestige. When the matter is closelyexamined
the urban proletariat, as well as the rural citi-
zenry, proves to be surprisingly god-fearing and
priest-ridden. No society has become so com-
pletely secularized as to liquidate entirely the be-
lief in transcendental ends and supernatural enti-
ties. Even in a secularized society some system
TheFunctions of Stratification
must exist for the integration of ultimate values,
for their ritualistic expression, and for the emo-
tional adjustments required by disappointment,
death, and disaster.
Government
Likereligion, government plays a unique and in-
dispensable part in society. But in contrast to reli-
gion, which provides integration in terms of sen-
timents, beliefs, and rituals, it organizes the
society in terms of law and authority. Further-
more, it orients the society to the actual rather
than the unseen world.
The main functions of government are, inter-
nally, the ultimate enforcement of norms, the fi-
nal arbitration of conflicting interests, and the
overall planning and direction of society; and ex-
ternally, the handling of war and diplomacy. To
carry out these functions it acts as the agent of the
entire people, enjoys a monopoly of force, and
controls all individuals within its territory.
Political action, by definition, implies author-
ity. An official can command because he has au-
thority, and the citizen must obey because he is
subject to that authority. For this reason stratifi-
cation is inherent in the nature of political rela-
tionships.
So clear is the power embodied in political po-
sition that political inequality is sometimes
thought to comprise all inequality. But it can be
shown that there are other bases of stratification,
that the following controls operate in practice to
keep political power from becoming complete:
(a) The fact that the actual holders of political of-
fice, and especially those determining top policy
must necessarily be few in number compared to
the total population. (b) The fact that the rulers
represent the interest of the group rather than of
themselves, and are therefore restricted in their
behavior by rules and mores designed to enforce
this limitation of interest. (c) The fact that the
holder of political office has his authority by vir-
tue of his office and nothing else, and therefore
any special knowledge, talent, or capacity he may
claim is purely incidental, so that he often has to
depend upon others for technical assistance.
In view of these limiting factors, it is not
strange that the rulers often have less power and
43
prestige than a literal enumeration of their formal
rights would lead one to expect.
Wealth, Property, and Labor
Every position that secures for its incumbent a
livelihood is, by definition, economically re-
warded. For this reason there is an economic as-
pect to those positions (e.g. political and reli-
gious) the main function of which is not
economic. It therefore becomes convenient for
the society to use unequal economic returns as a
principal means of controlling the entrance of
persons into positions and stimulating the per-
formance of their duties. The amount of the eco-
nomic return therefore becomes one of the main
indices of social status.
It should be stressed, however, that a position
does not bring power and prestige because it
draws a high income. Rather, it draws a high in-
come because it is functionally important and the
available personnel is for one reason or another
scarce. It is therefore superficial and erroneous to
regard high income as the cause of a man's power
and prestige, just as it is erroneous to think that a
man's fever is the cause of his disease"
The economic source of power and prestige is
not income primarily, but the ownership of capi-
tal goods (including patents, good will, and pro-
fessional reputation). Such ownership should be
distinguished from the possession of consumers'
goods, which is an index rather than a cause of
social standing. In other words, the ownership of
producers' goods is properly speaking, a source of
income like other positions, the income itself re-
maining an index. Evenin situations where social
values are widely commercialized and earnings
are the readiest method of judging social posi-
tion, income does not confer prestige on a posi-
tion so much as it induces people to compete for
the position. It is true that a man who has a high
income as a result of one position may find this
money helpful in climbing into another position
as well, but this again reflects the effect of his ini-
tial, economically advantageous status, which ex-
ercises its influence through the medium of
money.
In a system of private property in productive
enterprise, an income above what an individual
44
spends can give rise to possession of capital
wealth. Presumably such possession is a reward
for the proper management of one's finances
originally and of the productive enterprise later.
But as social differentiation becomes highly ad-
vanced and yet the institution of inheritance per-
sists, the phenomenon of pure ownership, and re-
ward for pure ownership, emerges. In such a case
it is difficult to prove that the position is func-
tionally important or that the scarcity involved is
anything other than extrinsic and accidental. It is
for this reason, doubtless, that the institution of
private property in productive goods becomes
more subject to criticism as social development
proceeds toward industrialization. It is only this
pure, that is, strictly legal and functionless owner-
ship, however, that is open to attack; for some
form of active ownership, whether private or
public, is indispensable.
One kind of ownership of production goods
consists in rights over the labor of others. The
most extremely concentrated and exclusive of
such rights are found in slavery, but the essential
principle remains in serfdom, peonage,
encomienda, and indenture. Naturally this kind
of ownership has the greatest significance for
stratification, because it necessarilyentails an un-
equal relationship.
But property in capital goods inevitably intro-
duces a compulsive element even into the nomi-
nally free contractual relationship. Indeed, in
some respects the authority of the contractual
employer is greater than that of the feudal land-
lord, inasmuch as the latter is more limited by
traditional reciprocities. Even the classical eco-
nomics recognized that competitors would fare
unequally, but it did not pursue this fact to its
necessary conclusion that, however it might be
acquired, unequal control of goods and services
must give unequal advantage to the parties to a
contract.
Technical Knowledge
The function of finding means to single goals,
without any concern with the choice between
goals, is the exclusively technical sphere. The ex-
planation of why positions requiring great tech-
nical skill receivefairlyhigh rewards is easy to see,
II I Forms and Sources ofStratification
for it is the simplest case of the rewards being so
distributed as to draw talent and motivate train-
ing. Why they seldom if ever receive the highest
rewards is also clear: the importance of technical
knowledge from a societal point of view is never
so great as the integration of goals, which takes
place on the religious, political, and economic
levels. Since the technological level is concerned
solely with means, a purely technical position
must ultimately be subordinate to other positions
that are religious, political, or economic in char-
acter.
Nevertheless, the distinction between expert
and layman in any social order is fundamental,
and cannot be entirely reduced to other terms.
Methods of recruitment, as well as of reward,
sometimes lead to the erroneous interpretation
that technical positions are economically deter-
mined. Actually, however, the acquisition of
knowledge and skill cannot be accomplished by
purchase, although the opportunity to learn may
be. The control of the avenues of training may in-
here as a sort of property right in certain families
or classes, giving them power and prestige in con-
sequence. Such a situation adds an artificial scar-
city to the natural scarcity of skills and talents. On
the other hand, it is possible for an opposite situ-
ation to arise. The rewards of technical position
may be so great that a condition of excesssupply
is created, leading to at least temporary devalua-
tion of the rewards. Thus "unemployment in the
learned professions" may result in a debasement
of the prestige of those positions. Such adjust-
ments and readjustments are constantly occur-
ring in changing societies; and it is alwayswell to
bear in mind that the efficiency of a stratified
structure maybe affected by the modes of recruit-
ment for positions. The social order itself, how-
ever, sets limits to the inflation or deflation of the
prestige of experts: an over-supply tends to de-
base the rewards and discourage recruitment or
produce revolution, whereas an under-supply
tends to increase the rewards or weaken the soci-
ety in competition with other societies.
Particular systems of stratification show a wide
range with respect to the exact position of techni-
cally competent persons. This range is perhaps
most evident in the degree of specialization. Ex-
treme division of labor tends to create many spe-
TheFunctions of Stratification
cialists without high prestige since the training is
short and the required native capacity relatively
small. On the other hand it also tends to accentu-
ate the high position of the true experts-scien-
tists, engineers, and administrators-by increas-
ing their authority relative to other functionally
important positions. But the idea of a techno-
cratic social order or a government or priesthood
of engineers or social scientists neglects the limi-
tations of knowledge and skills as a basis for per-
forming social functions. To the extent that the
social structure is truly specialized the prestige of
the technical person must also be circumscribed.
Variation in Stratified Systems
The generalized principles of stratification here
suggested form a necessary preliminary to a con-
sideration of types of stratified systems, because it
is in terms of these principles that the types must
be described. This can be seen by trying to delin-
eate types according to certain modes of varia-
tion. For instance, some of the most important
modes (together with the polar types in terms of
them) seem to be as follows:
(a) The Degree of Specialization. The degree of
specialization affects the fineness and multiplicity
of the gradations in power and prestige. It also in-
fluences the extent to which particular functions
may be emphasized in the invidious system, since
a givenfunction cannot receivemuch emphasis in
the hierarchy until it has achievedstructural sepa-
ration from the other functions. Finally, the
amount of specialization influences the bases of
selection. Polar types: Specialized, Unspecialized.
(b) The Nature of the Functional Emphasis. In
general when emphasis is put on sacred matters, a
rigidity is introduced that tends to limit special-
ization and hence the development of technology.
In addition, a brake is placed on social mobility,
and on the development of bureaucracy. When
the preoccupation with the sacred is withdrawn,
leaving greater scope for purely secular preoccu-
pations, a great development, and rise in status,
of economic and technological positions seem-
ingly takes place. Curiously, a concomitant rise in
45
political position is not likely,because it has usu-
ally been allied with the religious and stands to
gain little by the decline of the latter. It is also pos-
sible for a society to emphasize family func-
tions-as in relatively undifferentiated societies
where high mortality requires high fertility and
kinship forms the main basis of social organiza-
tion. Main types: Familistic, Authoritarian (Theo-
craticor sacred, and Totalitarian or secular), Cap-
italistic.
(c) The Magnitude of Invidious Differ-
ences. What may be called the amount of social
distance between positions, taking into account
the entire scale, is something that should lend it-
self to quantitative measurement Considerable
differences apparently exist between different
societies in this regard, and also between parts of
the same society. Polar types: Equalitarian,
Inequalitarian.
(d) The Degree of Opportunity. The familiar
question of the amount of mobility is different
from the question of the comparative equality or
inequality of rewards posed above, because the
two criteria may vary independently up to a
point. For instance, the tremendous divergences
in monetary income in the United States are far
greater than those found in primitive societies,
yet the equality of opportunity to move from one
rung to the other in the social scale may also be
greater in the United States than in a hereditary
tribal kingdom. Polar types: Mobile (open), Im-
mobile (closed).
(e) The Degree of Stratum Solidarity. Again,
the degree of "class solidarity" (or the presence of
specific organizations to promote class interests)
may vary to some extent independently of the
other criteria, and hence is an important princi-
ple in classifying systems of stratification. Polar
types: Class organized, Class unorganized.
External Conditions
What state any particular system of stratification
is in with reference to each of these modes of vari-
ation depends on two things: (1) its state with ref-
MELVIN M. TUMIN
Some Principles of Stratification: A Critical Analysis
The Dysfunctions of Stratification
46
erence to the other ranges of variation, and (2)
the conditions outside the system of stratification
which nevertheless influence that system. Among
the latter are the following:
(a) The Stage of Cultural Development. As the
cultural heritage grows, increased specialization
becomes necessary, which in turn contributes to
the enhancement of mobility, a decline of stratum
solidarity, and a change of functional emphasis.
(b) Situation with Respect to Other Societies.
The presence or absence of open conflict with
other societies, of free trade relations or cultural
diffusion, all influence the class structure to some
extent. A chronic state of warfare tends to place
emphasis upon the military functions, especially
when the opponents are more or less equal. Free
trade, on the other hand, strengthens the hand of
the trader at the expense of the warrior and priest.
Free movement of ideas generally has an equali-
tarian effect. Migration and conquest create spe-
cial circumstances.
(c) Size of the Society. A small society limits the
degree to which functional specialization can go,
the degree of segregation of different strata, and
the magnitude of inequality.
Composite Types
Much of the literature on stratification has at-
tempted to classify concrete systems into a certain
number of types. This task is deceptively simple,
however, and should come at the end of an analy-
sis of elements and principles, rather than at the
beginning. If the preceding discussion has any va-
lidity, it indicates that there are a number of
modes of variation between different systems,
and that anyone system is a composite of the so-
ciety's status with reference to all these modes of
variation. The danger of trying to classify whole
societies under such rubrics as caste, feudal, or
open class is that one or two criteria are selected
1\ I Forms and Sources 01 Stratification
and others ignored, the result being an unsatisfac-
tory solution to the problem posed. The present
discussion has been offered as a possible ap-
proach to the more systematic classification of
composite types.
Notes
1. Kingsley Davis,"AConceptual Analysis of Stratifi-
cation;' AmericanSociological Review. 7:3
09-321,
June,
'942.
2. The writers regret (and beg indulgence) that the
present essay, a condensation of a longer study, covers
so much in such short space that adequate evidence
and qualification cannot be given and that as a result
what is actuallyverytentative is presented in an unfor-
tunately dogmatic manner.
3. Unfortunately, functional importance is difficult
to establish.To usethe position's prestigeto establishit,
as is often unconsciouslydone, constitutes circular rea-
soning from our point of view.There are, however, two
independent clues: (a) the degreeto whicha position is
functionally unique, there being no other positions
that can perform the same function satisfactorily; (b)
the degree to which other positions are dependent on
the one in question. Both clues are best exemplified in
organized systems of positions built around one major
function. Thus, in most complexsocietiesthe religious,
political, economic, and educational functions are han-
dled by distinct structures not easily interchangeable.
In addition, each structure possesses many different
positions, some clearly dependent on, if not subordi-
nate to, others. In sum, when an institutional nucleus
becomes differentiated around one main function, and
at the same time organizesa largeportion of the popu-
lation into its relationships, the keypositions in it are of
the highest functional importance. The absenceof such
specializationdoes not provefunctional unimportance,
for the whole society may be relatively unspecialized;
but it is safe to assume that the more important func-
tions receivethe first and cleareststructural differentia-
tion.
4. The symbolic rather than intrinsic role of income
in social stratification has been succinctlysummarized
by Talcott Parsons, "An Analytical Approach to the
Theory of SocialStratification;' American Journal 0[50-
ciology. 45:841-862, May, 1940.
The fact of in human society is
marked by. ItS ubiquity and its antiquity. Every
known society, past and present, distributes its
scarce and demanded goods and services un-
And there are attached to the positions
which command unequal amounts of such goods
a.nd highly morally-toned evalua-
nons of Importance for the society.
. The ub.,qUlty.and the antiquity of such inequal-
ity has given nse to the assumption that there
must .be something both inevitable and positively
functional about such social arrangements.
. the truth or falsity of such an assump-
non IS. a strategic question for any general theory
ofSOCial organization. It is therefore most curious
that basic premises and implications of the as-
sumption have only been most casually explored
by American sociologists.
. The most systematic treatment is to be found
in .the well-known article by Kingsley Davis and
entitled "Some Principles of
Stratification, More than twelve years have
passed since its publication, and though it is one
0: the very few treatments of stratification on a
level of generalization, it is difficult to locate
a single analysis of its reasoning. It will
be the concern of this paper to present
the beginnings of such an analysis.
The central argument advanced by Davis and
Moore be stated in a number of sequential
proposrtions, as follows:
(1) positions in any society are func-
tionally more important than others and
require special skills for their
(2) Only a limited number of individuals in
society have the talents which can be
into the skills appropriate to these
posrtions.
(3) The conversion of talents into skills in-
v?lves a training period during which sac-
nfices of one or another are made by
those undergoing the training.
(4) In order to induce the talented persons to
sacrifices and acquire the
their future positions must carry
an value in the form of differ-
ential, i.e., privileged and disproportionate
access to the scarce and desired rewards
which the society has to offer.'
(5) scarce and desired goods consist of
the rights and perquisites attached to or
built into, the positions, and can be classi-
fied into those things which contribute to
(a) sustenance and comfort, (b) humor
and (c) self-respect and ego ex-
pansion,
(6) This differential access to the basic rewards
of the society has as a consequence the dif-
ferentiation of the prestige and esteem
various strata acquire. This may be
said, alo.ng t.he rights and perquisites,
to constitute institutionalized social in-
equality, i.e., stratification.
(7) Therefore, social inequality among differ-
ent strata in the amounts of scarce and de-
sired goods, the amounts of prestige
and esteem which they receive, is both
47