On the Concept of T ypes

A n article by J. A. FORD with discussion by J. H. STEWARD

T he Type Concept Revisited1

O E V E R A L years ago, K luckhohn (1939) upbraided anthropologists in gen-
eral a n d archeologists in p articular for failure to examine critically the
assu m p tion s and concepts which lie a t the foundations of their methodologies.
P e rh a p s this well justiñ ed censure has p rom pted the healthy introspection th a t
has developed in the p ast decade an d resulted in valuable papers such as those
by Rouse (1939), Krieger (1944), Brew (1946), T a y lo r (1948), an d Ehrich
(1950).
As soon a s s tu d e n ts o f cultural phenomena cease to be satisñed with com-
parisons oí mere qualities of cultural traits an d begin also to fcreat their da ta
q u an tita tiv e ly , it becomes a p p a re n t th a t the basic conceptual tool of cultural
research is th a t of the type. To the present it is the archeologists who have
been m ost concerned with the formulation and use of cultural types, b u t this
h a rd ly redounds to the credit of this branch of the profession. Archeologists
have been forced into this position by the necessity for reconstructing cultural
histories from a v ery limited range of cultural material. Although the term has
been used indiscriminately, in practice the typological concept has been
thou gh tfu lly applied alm ost entirely to ceramics. The principies are the same,
however, for all o ther aspects of culture, and we m ay expect to see it more
widely used as sufficient evidence accumulates to make it possible and neces-
sary.
T o utilize the concept of type efficiently, it is very necessary th a t the cul­
tu ra l s tu d e n t have a clear idea of w hat a type is, how it is defined, and w hat
purposes it m a y serve. A t present there seems to be some confusion. The de­
b a te seems to center aro u n d the question of the “ realitv” of cultural tvpes; a
debate which is very similar to th a t carried on by the biologists for a num ber
of years in regard to the significance of the species concept. To state it clearly,
the question m a y be p u t this w ay: “ Do cultural types exist in the phenomena
so th a t they m a y be discovered by a capable typologist?” This is an im po rtan t
question for the answer not only determines how investigators may proceed
in identifying types, b u t it also determines how types m a y be employed in
solving cultural problems.
B o th Rouse (1939) an d Krieger (1944) have given excellent discussion of
the application of the concept of type b u t have failed to clarify this debated
point. N e ith e r am I entirely satisñed with the sta tem e n t in Phillips, Ford and
Griffin (1951:61-64). Recently the question has again been brought up as a
result of an article b y A. C. Spaulding (1953) which describes a method for
discovering cultural types b y statistical methods. This discussion takes for
g ra n te d the assum ption th a t types do exist in culture and m ay be discovered
by co m p etent methodologies. T his I doubt.

• i ' ñ ,r \ \ ■5 ’ ‘‘

' V í r " >i r ~

. , ■- -*
 [p o r d ] On the Concept of Types 43

Perhaps it will clarify th e problem to say a word abo ut the history of the
type concept, for the purposes of classification of archeological material have
'y J- H. STEWARD undergone a change beginning in this country during the second an d third
decades of this century. Initially archeological classifications were made for
the purpose of describing collections, a n d the smallest divisions of the Ítems
were frequently called types. These groupings were defined without reference
ropblogists in gen-
to the temporal a n d spatial coordinates of culture history. Where chronological
ine; criticall) ■ the
Information is lacking such descriptive classifications are the only sort th a t can
eir methodologies.
be made an d are extrem ely useful. A good example of such a classification is
i»tr¿>spection th a t
S. K. L o th ro p ’s (1926) analysis of p o tte ry collections from Costa Rica and
pers; such as those Nicaragua.
948).. a n d Ehrich The classifying of ceramics into type groupings th a t are designed to serve
as measuring devices for culture history began in the southwestern United
|tisfifed w ith com-
States and is now s ta n d a r d practice am ong American archeologists. Descrip­
p tr e a t their da ta tive system atization is sub ordinated to the necessity for emphasizing spatial
kl toW of cultural and temporal change in the material. Perhaps it is unfortu nate th a t the word
jlogisis who have
“ ty p e ” has been retaine.d for this new function, because to some it seems to
|U types, b u t this carry a connotation of its earlier descriptive usage. Krieger (1944:272) has
|n . Archeologists stated the current purpose of formulating tvpes in the following words:
Itructing cultural
Thus the purpose of a type in archaeology must be to provide an organizational
|igh the term has
tool which will enable the investigator to group specimens into bodies which have

Í
cepjt has been
demonstrable historical meaning Í7i terms of behavior patterns. Any group which may be
a aije the same, labelled a “ type” must embrace material which can be shown to consist of individual
to kee it more variations in the execution of a definite constructional idea; likewise, the dividing lines
¡b le ¡and neces- between a series of types must be based upon demonstrable historical factors, not, as is
often the case, upon the inclinations of the analyst or the niceties of descriptive orderli-

Í
y thjat the cul- ness.
ned, and what
Spaulding (1953:305) seems to agree th a t to be useful each type must have
usion. T h e de-
historical significance: “ H istorical relevance in this view is essentially derived
iltura! types; a
from the typological analysis; a properly established type is the result of sound
| for a num ber
inferences concerning th e cu sto m arv behavior of the makers of the artifacts
jtate it clearly,
and cannot fail to have historical meaning.” I certainly am in agreement with
(ie phenomena
1 both these authors th a t to be useful, each type m ust have a limited range in
¡an im p o rta n t
time and space and th u s have historical significance.
f m a y proceed
The discussion t h a t follows will retrace some of the same argum ents set
| employed in
forth by Rouse and Krieger b u t will consider typology from a slightly different
angle. Instead of emphasizing the-problem from the point of view of archeo­

!
disci|ission of
logical specimens, I shall examine the concept as it would apply to a living
th i s d e b a te d
3s, Ford and culture. Further, to m ak e the task easier and to a tte m p t to clarify basic prob-
lems which the typologist m u st face, this will be fictitious culture history which
íght lup as a
has not beén subjected to the complicating factors th a t operate in all actual
m ethod for
histories. These factors are barriers to diffusion such as uneven pupulation
>n t^kes for
: disfcovered distribution, na tu ra l obstacles to communication, political and linguistic
boundaries, or boundaries between competing cultural items of different geo-
graphic origin. N eith er will it be subjected to the forces th a t speed and retard
44 A m erican Anlhropologist [56 , 1954

cultural change— wars, epidemics, alien cultures with high prestige, or adver-
tising by influential innovators. E ach culture bearer has been the normal minor
innovator th a t has borne the responsibility for most of the change th a t has
taken place in culture histories.
T he fairy tale which follows is the sort of “ strip p e d ” description of phe-
nomena which has proved very useful in more m a tu re fields of Science, such as
physics. E very physical “ Law” states th a t if certain modifying circumstances
were nullified such an d such would happen. In experience the modifying cir­
cumstances are always présent a n d events never conform exactlv to the
■'Law.” This, then, is my excuse for introducing the Gamma-gamma, people of
the Island of Gam ma, situated in the curious sea of Zeta.

A C U L T U R E IS A C L A S S IF IC A T O R Y D E V IC E

W ith no intention to disparage the work of fellow anthropologists, it may

be said t h a t the synchronous view of the ethnologist is the most simple way
to consider cultural phenomena. W hen an ethnologist ñ rst arrives among the
G am m a-gam m a of the Island of G am m a, their culture will impress him as a
confused co nglom eraron of absurdities. T he G am m a-gam m a will do strange,
unreasonable things and on m any occasions will app ear to be lacking in com-
mon horse sense— an impression th a t has been shared by every tourist who
has come into c ontact with people h aving a culture different from his own.
As the more-or-less impartial ethnologist becomes b e tte r acquainted and
begins to acquire something of the po int of view of the-G am m a-gam m a, social
actions an d cultural objects begin to fall into classes. I t will be discovered that
these classes are well organized to solve the problema th a t confront this group
of h um a n animals: procuring food, providing shelter and protection from ene-
mies, regulating m ating and other social relatiuns, and magical techniques that
affect otherwise uncontrollable forces such as diseases and the weather. There
a^e p a tte rn e d ways of dancing, of constructing a canoe, of clothing and deco-
rating the body, etc. In addition, if the basic premises of Gamma-gamma
th o ugh t are accepted, m any of these cultural categories have a logical, appar-
ently inevitable, relation to one a n o th e r and these relations are cross-ties th a t
reinforce and stabilize the entire cultural structure. Certain dances are neces-
sarv as a preliminary to catching fish; a m an cannot m arry until he has killed
an enemy— hu m an or shark— an d has been tattoo ed ; houses are the property
of women because they build them ; children belong to the m o th er’s family
for where is the child who can be certain of his father?
This com partm entalization and order are necessary and will be found in all
other cultures. To add to the definitions recently listed by Kroeber and
Kluckhohn (1952), it can also be said th a t culture is an organized system for
handling hum an an d social problems. However, different segments of a culture
will vary as to the range of variability which is p erm itte d as acceptable be­
havior. T he Gam m a-gam m a group h as'very strict rules as to how a man may
address his mother-in-law: he must face away to avoid seeing her and preface
all remarks with polite formal phrases— to do otherwise would cause great
 [56, 1954 On the Concept of Types
prestjige, or adver- scandal and w hat else no one knows for it has never been tried. However, there
the normal minor are a num ber of perfectly good ways to m ake an adze. Virtually any hard stone
change th a t has will serve as a blade, four varieties of hafting are used, a n d there are six shapes
of handles. In addition, a m an takes some pride in carving the handle in an
pscrif)tion of phe- original fashion, as different from those of his fellows as possible. Still, any
|f scjence, such as ethnologist acquainted with the material culture of this región can recognize a
ng cjrcumstances tw entieth-century G am m a-gam m a adze a t a glance. Despite the: fact th a t it
he modifying.cir- permits and even appears to encourage variability, this cultural trait is a
exactly to the classificatory device similar to the mother-in-law taboo and has wider but still
«gamma people of rigid limits. T he variation follows p a tte rn s and these people h av en’t thought
of turning the blade' arou nd and making a h a tc h e t of the tool.
I t is this inherent order in culture of which archeologists must be aware
when they begin the search for types for this is the framework within which
pologists, it may
the typology m ust be constructed. This is certainly the order th a t will be re-
fiost; simple way
vealed by applying statistical devices to the ceramics of prehistoric dwelling
prives among the
sites as recently ad vocated by Spaulding (1953). However, this order does not
jnpréss him as a
provide the historically significant grouping of traits which the archeologist
hvill do strange,
must have to measure culture history.
lacking in com-
ery tourist who T H E E T H N O L O G IS T ’S VIEYV O F A C U L T U R A L T R A I T
’ rom his own. T he G am m a-gam m a have each aspect of their culture well compartmental-
acquainted and ized: p o tte ry íood-serving vessels have a limited range of shapes and decora-
i-gamma, social tions; w ater bottles have their appropriate range; an d the containers in which
discovered tha t the mild alcoholic drink is fermented have their range. However, the actual
íro n t this group specimens th a t fa re m anufactured for these various purposes are by no means
jbtioti from ene- identical duplicates such as would be turned out by a machine. Instead, each
lechniques th a t vessel is recognizably different from every oth er vessel in its class. As the eth­
iweather. There nologist studies the p ottefy, and other aspeets of'the culture, he will observe
Jhing and deco- th a t the variation in actual artifact tends to cluster ab o u t a mean, which he
lamma-gamrna can then visualize as the central theme of the type.
logical, appar- T he ethnologist cannot rely upon the culture bearers to define this central
cross-ties th a t theme. T h e y m ay or m a y not be aware of it, or m ay have rationalizations in

t ees are neces-

t he has killed
e the p roperty
regard to it which are a t considerable variance with actual practice, as Dr.
K in sey ’s study of male sexual practices has dem on strated for our own culture.
A statistical average m ust be arrived at, either by actual counting or by esti-
Lther's
i
familv mating. If desirable, the rationalizations m a y be considered a p a rt for they are
also cultural features and are subject to the same kind of analysis as actions.
¡)e found in all The cultural trait, then, is an abstractio n made by the ethnologist and
Kroeber and derivedfrom the cultural activity. I t has a m ean and a range of variation. This
ed system for range of variation m ay be visualized as a sc a tte r diagram— a three-dimensional
ts of:a culture sc a tte r diagram similar to a swarm of bees clustering a b o u t the queen might
cceptable be- better represent the situation, b u t there are limitations to the printed page
¡v a man may and a two-dimensional diagram will have to serve. In Figure 1, I have at-
r a nd preface tem pted to represent the variation in houses th a t was observed among the
cause great G am m a-gam m a on the Island of G am m a in 1940. As the diagram shows, the
46 A m erican Anthropologist [56, 1954

m ajority of houses were medium-sized rectangular structures a b o u t 4 by 6 are all

m eters an d 5 m eters high, placed on low piers above the dam p ground, and there
had gabled roofs, one room, a n d one door. Variations from this norm are ob- exami
served in several directions. Houses illustrated tow ard the right of the diagram, the di
mostíy occupied by older people, were on high stilts, and one is in a tree. They limits
room.
b a th r
lógica
a me;
consii

m ent
ideal
conc:
aspei
unit
th at
men
caus
dwe
The
and
be r
foui
legi

F ig . 1. A diagram illu stra tin g th e freq u e n cy m ean of a ty p e a t one p o in t in tim e an d space.

T h e sm all houses grouped in th e in n e r circle rep re sen t th e m ean. V ariatio n from th is m ean is illus­
tra te d in four d irections— a sim plification of th e v ariatio n th a t is íound a b o u t th e frequency m eans
o f,actu al types.

tend to be smaller th a n the average. T ow a rd the left side of the diagram, the
houses are larger and are on v ery low stilts, or are built on the ground. A few
have two rooms. V ariation tow ard the top of the diagram tends tow ard larger
size, and tow ard the b o tto m the houses are small, square, and the roofs ap-
proach the pyram idal in shape.
This description is an obvious simplification. As the diagram shows, there
 [56, 1954 On the Concept of Types 47
cturps a b o u t 4 by 6 are all sorts of variations between the four poles described and, in addition,
e d am p ground, and there are other variables which could also serve as poles in this diagram. For
xa this norm are ob- example, some buildings are roofed w ith the white palm fronds an d on others
ríght of the diagram,
the dark gray kilea grass is used. Still, these combinations have definite
ane is in a tree.! T hey
limits of variation. N one of the houses has more than one living and sleeping
room, all are constructed of bam boo an d thatch, and no one has introduced
bathrooms such as are observed in the local mission buildings. To the ethno-
logical observer it is quite clear th a t there is a G am m a-gam m a house type with
a mean and range of v a ria tion as ju st described. In Figure 1, w hat m ay be
considered the m ean of the type lies w ithin the inner circle.

T H E E T H N O G R A P H IC T Y P E IS F O R M E D B Y T H E O B S E R V E R A T
A C H O S E N L E V E L O F A B S T R A C T IO N

The dwellings of the G am m a-gam m a a t first glance offer a convenient seg-

ment of their culture composed of tangible elements and would seem to be
ideal for the purpose of measuring. Upon closer examination, the a pp a re n t
concreteness of this category can.be broken down in two directions, for this
aspect oí the culture is p a r t of a n in tegrated whole an d became a measurable
unit merely because a tte n tio n was focused upon it. First, it m ust be recalled
th a t these buildings are cultural p ro du cts— not the culture.' These arrange-
ments of wood, bamboo, an d grass are of interest to the ethnologist solely be-
cause they ¿Ilústrate the aborigine’s ideas as to the proper ways to construct
dwellings. T he cultural concept “ house” can be broken down into elements.
There are a range of m ethods to anchor piers, to arrange plates, to lash rafters,
and a t least four s ta n d a r d m ethods of thatching. Each of these elements can
be measured in the same w ay as the entire houses have been and each will be
found to have a frequency m ean and range of variation. “ H ouse” m ay quite
Jegitimately be considered as a cultural complex rather than as a unit.2
On the other hand, the concepts dictating the proper ways to build a house
are not isolated in the culture. For one thing, they are intimately connected
with the form of the family. These people are monogamous and married chil-
dren set themselves up in separate establishments. There is never need for
more than one living room, ñor is large size necessary. In tura, the single-room
small houses tend to reinforce this p a tte rn of family life. The house, then, might
legitimately be considered as one element of the 1940 G am m a-gam m a family
|nt in;!time an d space, type.
rom th is m ean is illus-
I t is evident th a t “ cultural typ es” are abstracted on different levels of
f the fpequency meims
apparent complexity by the observer.3 One level is no m o re - “ real” than
' i f another. W hat the classifier m u st do is to select a level which will serve the
the diagram, the purposes in view. If the objective is a comparison of religións, the stu den t will
fie ground. A few set up religious types; if it is concerned with priestly paraphernalia, the types
ds tpw ard larger will be formed of cultural traits which are mere elements for the preceding
n d the roofs ap- purpose. T he cultural scientist m ust be aware of this necessity and not allow
chance focalization of in terest to provide categories th a t are accepted as im­
;am shows, there mutable units.
48 A m erican Anlhropologist [56, 1954
T H E E T H N O G R A P H IC T Y P E IS A B S T R A C T E D B Y T H E O B S E R V E R
A T O N E P O IN T IN SPA CE

So long as the ethnologist stays am on g the Gam m a-gam m a on the Island.

of Gamma, the house typ e described above appears to íorm a satisfactory
unit. I t seems to be a n a tu ra l división of the culture. However, in the surround-
ing territory live people with the same general cultural tradition as the in-
h a b ita n ts of Gamma. A fter the ethnologist finishes his preliminary survey of
G am m a an d begins to visit their neighbors, he will discover th a t there is'
an other reason why the house type which he has described for the Gamma-
gam m a is n ot a natu ral cultural unit.
In Figure 2 is illustrated the frequency distributions of dwellings on the
islands th a t lie about the Isle of G am m a. This is a very simplified diagram. On
each island the house in the center represents the mean as illustrated in Figure
1; the four buildings arran ged ab o u t each-mean represent the range of varia-

G am m o)*

F ig . 2. D iagram illu stratin g tr a it v a ria tio n in geographical space. T h e Islan d of G am m a oc-

cupies th e center. T h e frequency of th e G am m a-g am m a house ty p e is in th e cen ter of this island
a n d th e less n um erous variatio n s are grouped a ro u n d it. O n neighboring islands m ean and range
are sim ilarly in d icated .
 [56, 1954
*E O B S E R V E R
gámína on the Island
t> form a satisfactory
íever, in the surround-
tradition as the in-
reliminary survey of
¡SCOver th a t thexe is
>ed for the Gamma-
of dwellings on the
^nplified diagram. On
ted in Figure
range of varia-
of G am m a oc~
of th is island
|a n d s m e a n and range
Ford] On the Concept of Types 49
tion. I t becomes a p p a re n t th a t the G am m a-gam m a house type, illustrated in
Figure 1, is n o t the cohesive cultural type which it appeared to be. T he variants
from the m ean have to be assigned to house types typical of the neighboring
peoples north, south, east* an d west. As a m a tte r of fact, this diagram shows
th a t the p olarity of Figure 1 was n ot correct. All the black-roofed houses are
related to a type th a t centers to the eastward. Very few examples need be left
to be classed in the G am m a-gam m a type h o u s e .'
Figure 2 illustrates the p oint th a t each locality will have a distinctive mean
and a range a b o u t t h a t mean which tends toward the m eans of surrounding
culture. However, Figure 2 is an unsatisfactory diagram in th a t the geographi-
cal separation of the islands has created nodes in the p a tte rn of distribution.
If the landscape had been undivided, the geographical variation would be a
more gradual function of space, similar to th a t shown in Figure 3. Although
this la tte r figure is designed to demónstrate th e .n a tu re of change with time, it
will serve equally well for this discussion of space-change. F o r this purpose it
will be considered th a t each building shown represents a local type. Variation
about the mean in each locality is not shown. T he building near the upper
center of the figure, ju s t above the hurrying female in a grass skirt, will repre­
sent the m ean type a t G am m a-gam m a. The gradualness of the, change in
means in all directions becomes apparent.
Lest the reader suspect th a t this description is puré fiction, he is referred
to an article by Wilhelm Milke which summarizes several illustrations th a t
qualitative differentiation in culture is a function of distance.4 For an illustra-
tion th a t the popularity of specific cultural categories is also a function of geo-
graphic space, see Phillips st ai. (1951: Figs. 6-12) an d F ord (1952).
Setting aside the fictional Gam ma-gamma for the m om ent, in actual dis-
tributions of cultural items change in form is accelerated by natu ral, politicai,
and linguistic barriers, or a t the zones where competing cultural items of differ­
ent origins meet. For several reasons these barriers can no t be depended upon
to furnish lim itations to the spatial aspect of the variations th a t m ay be in-
cluded in a cultural type. First, there m ay be no such barriers operating on the
selected cultural item in the región under study— it is certainly not legitímate
to assume th a t there were before their eífect can be m easured by the typology.
Second, the eífect of such barriers is often less th a n m ight be imagined. With
the exception of impassable terrain, the eífect of a barrier is usually to produce
a more or less b road zone in which the rate of change with geographic space is
accelerated.
I t follows, then, th a t the particular locality where an archeological collec-
tion chances to be m ade will be one of the factors th a t determines the mean
and the range of v ariation th a t are dem onstrated in a n y p a rticular tradition in
the culture th a t is being studied. On the same time level, the cordmarked
p ottery from a village site in northern Illinois is different from th a t on a site
in Southern Illinois. If the archeologist has only these two collections to study
and is not conscious of the n ature of the problem, separate types may be
established” and considered as realities, unconscious of the favor performed
SO Am erican Anlhropologist [56, 1954
by the chance geographic separation of samples, However, if additional col­
lections, all of the same date, are available to fill in the intervening space, then
the problem usually becomes .the difficult one of fixing boundaries in a con-
tinu um which Phillips has described (Phillips et al. 1951:66-68),

Gamma-
gam m a

F i g .'3 . T h is diagram will serve tw o p u rposes. F irs t, it will rep re sen t geographical d istrib u -
tio n of v ariatio n a n d for th is p u rp o se each h ouse rep re sen ts a tr a it m ean. L ocation of th e G am m a-
g am m a m ean is show n a n d p o sitio n of th e houses rep re sen ts geographical location. Second, this
will serve as a chronological d iag ram . F o r th is p u rp o se tim e is th e v ertical ordinate of th e figure
an d decades are in d icated on th e rig h t-h a n d side. V ariatio n is show n horizontally w ith frequency
m eans in th e cen ter of th e d ia g ra m .
■ T [56, 1954
P w e v e r , if additional col-
theiintervening space, then
íxin¿ boundaries in a con-
1951:66-68).
!|sent g eo g rap h ical d istrib u -
p . L o catio n of th e Gamma--
:ucal locatio n . Second, this
tical (jrd in ate of th e figure
p rizo n tally w ith frequency
j,oRD] On the Concept of Types 51
T H E T Y P E IS A B S T R A C T E D B Y T H E O B S E R V E R A T A
P O IN T I N T IM E
The ethnográphic view of a culturo resembles a snapshot taken in the
middle of a race for it is a static view of a very fluid process. Stretching back
in time from each cultural element described and measured by the ethnogra-
pher there is a long history which m u st be traced if we wish to know why the
tra it assumes its particular form. For cultural traits th a t did not find expres-
sion in durable form, this is impossible; it can be done, however, for enough
streams of thou gh t to dem ó n strate the principie beyond reasonable doubt.
As illustration, again consider the m ythical Gamma-gamma. In the 1940
static diagram, Figure 1, house structures are shown varying four ways from
the mean of the type— already a simplification of the variation as explained
above. To give a temporal picture of variation, it will be necessary to simplify
still further and show only two directions of variation from the mean. This
has been done in Figure 3, which now will be used for the purpose for which it
was designed. From b o tto m to top this diagram represents the passage of
time. Decades are indicated on the rig ht-h an d side. House form variation is
shown horizontally and the frequency m ean forms are illustrated down the
center of the figure.
T he phenomena of cultural drift with the passage of time is so well known
to archeologists who have dealt w ith a dequate samples of material culture
representing appreciable time spans of culture history th a t it does not seem
necessary to elabórate the illustration. E v e n in modern Western culture, with
all of the acceleration of change th a t has developed, the well paid innovators
who control design of automobiles, architecture, and clothing have learned
th a t while minor innovations will sell new models, the buying public will
tolerate no m arked jum ps in the developm ent of stylistic patterns.
Figure 3 cannot fully illustrate the phenom ena of time change among the
houses of the Gam m a-gam m a for cióse inspection of these structures would
show th a t not only did the gross outline of the structures change, but similar
change was taking place in minor details such as systems for placing rafters,
lashing, the methods of thatching, etc. T he ethnologist’s view of this cultural
type in 1900 would have h a d the same order of mean and range as his 1940
view, b u t the types would have been recognizably different. A glance up and
down the time scale dem onstrates th a t there are no natural limits to temporal
change in this cultural element which m ay be utilized as type boundaries.
In actual culture histories there are instances of major innovations which
will cause one stream of cultural developm ent to be replaced by another. An
example is the addition of the gasoline m otor.to the buggy to make the horse-
less carriage. This is a different order of innovation from the numerous small
changes th a t have occurred in the design of wheeled personnel carriers from
the invention of the first ca rt to the rubber-tired buggy, or from Charles
D u r y e a ’s automobile of 1892 to the 1953 Cadillac. Such major innovations are
so rare th a t the archeologist c a n not depend on them to provide temporal
52 Am erican Anlhropologisl [56, 1954

limits for typology. T hey are of little use íor the working out of details of
culture history.
A b r u p t change m a y also be caused b y accidents, or profound shocks to the
culture. For example, m any Pacific island peoples have taken advantage of
a b and on ed m ilitary establishments to change their dwelling types entirely.
T hese are also relatively rare and typology based upon them would measure
c ultural change in great blocks, not in a n y detail.
To summarize the preceding discussion, there are four dimensions to the
cultural type of which the archeologist m ust be fully aware if intelligent use is
to be m ade of the concept. These are:
(1) The inherent organization th a t exists in culture at all times and places.
T h e cultural type will, to a greater or lesser degree, be a retiection of the
boundaries to one stream of ideas which the cultural bearers considered re-
lated. This requires an analysis of the consistency of association of features
which may, if necessary, be tested by statistical analysis.
(2) T he level of abstraction from the tightly interwoven cultural structure
a t which the typology is to be formulated. For archeologists this may be at the
level of the artifact, or, if desirable, features of artifacts m ay be utilized as
Rouse has done for ceramics in the W est Indies.
(3) T he cultural type will encompass variation due to cultural drift across
geographical space. T he ap parent m ean of the type is the function of the
locality a t which it is deñned.
(4) T he cultural type will include variation th a t occurred with the passage
of time. T he a p p a re n t mean of the type is a result of the particular point in the
.histo ry of the cultural stream a t which it is selected.
In most archeological research, chance has determ ined the form of the
typological structure to a great extent. T h e fact th a t Site X was- in a certain
locality and represented a certain short span of culture history has determined
the n a tu re of the cultural types defined there. Perm ittin g sampling chance to
determ ine typology operates verv well so long as the archeologist has onlv a
sp o tty sampling of the culture history. T yp es are easily separable and they
look natural. However, when the gaps are filled in so th a t the history m ay be
viewed as a continuum through time and across space, the naive typologist is
certain to run into serious difficulties. Overlapping of types will render the
typology a meaningless conglomeration. The artific’iality of the groupjngs must
be taken into consideration and type groupings consciously selected if a work-
able typology is to be developed.
T h e type concept as discussed in this paper is the working tool of the
cultural stu d e n t— the device which is used to examine the most minute frag-
m ents of culture which the student can grasp. This tool is designed for the
reconstruction of culture history in time an d space. This is the begin-
ning and not the end of the archeologists’ responsibility. After culture history
has been outlined various other methods of classification become possible
and m a y be designed to measure different facets of the culture history. This,
I think, is the place for classifications based on function as described by

working out of delails of
ts, or profound shocks to the
have taken advantage of
dwelling types entirely.
asejd upón them would measure
le ta i.
|theiie are four dimensions to the
b e fully aware if intelligent use is
I '
I n culture a t all times and places,
'fer, degree, be a reflection of the
í cu tural bearers considered re-
Iteney of association of features
ical analysis.
tly ifiterwoven cultural structure
f archeologists this may be a t the
¡ of ^rtifacts m ay be utilized as
lie s.!
atioiji due to cultural drift across
|h e type is the function of the
ji th ^ t occurred with the passage
¡ult Of the particular point in the
ílectfed.
jis determ ined the form of the
ic t that Site X was in a certain
f culture history has determined
¡Perjnitting sampling chance to
; as the archeologist has only a
are; easily separable and thev
ji in so th a t the history may be
js space, the naive typologist is
ppinlg of types will render the
ificiality of the groupings must
congciously selected if a work-
ier i^ the working tool of the
:xamjine the most minute frag-
i This tool is designed for the
i d space. This is the beg'in-
nsibílity. After culture history
plassification become possible
js ofíthe culture history. This,
Ion function as described bv
í ¡
f q rd ] On the Concept of Types
Steward in th e accom panying paper. For example, the functional classifica­
tion which G ordon Willey (1953) applied to the prehistoric settlem ent pat-
terns in Viru Valley, Perú, v e ry neatly clarifies the history of this aspect of
culture a n d perm its com parison with the growth of communities in other parts
of the world. However, th e necessary prelude to this s tu d y of Willey’s was the
strictly morphological classification of thousands of potsherds.
Ja m e s A. F o rd , Am erican M useum of N atural H istory
XOTES
1 I wish to th a n k . Alex K rieger, P hilip Phillips, G ordon W illey, Ju liá n Stuw ard an d Irv in g
Rouse for read in g th e m a n u sc rip t of th is p a p e r an d m aking a n urnber o í h e tp íu l com m en ts. T h e
title is a bow to th e la te C laren ce B . M oore who freq u e n tly revisited prolilic archeological sites.
■ T h is is co m p arab le to w h a t Irv in g R ouse a n d o th e rs h a v e d o n e w h en th e v h a v e u tilized
ceram ic tr a its as bases for co m parison. R ouse term ed such elem en ts “ m o d es.”
3 “ A p p a re n t co m p lex ity ,” for a ll th ese levels a re infinitely com plex a n d it is th e lim ita tio n oí
th e observer’s a b ility to p erceiv e difíerences th a t set th e lim its. E h ric h (1950:468-81) gives an able
d is c u s s io n o f th i s m a t t e r .
4 M ilk e 1949. T h e w ord “ Q u a n tita tiv e ” in M ilk e’s title refers to th e n u m b ers of item s in th e
com pared c u ltu re s w h ich are sim ilar to th e reference culture— n o t to re lativ e p o p u la rity .
R E F E R E N C E S CIXED
B reiv , J . O tis
1946 A rchaeologv of A lkali R idge, so u th e astern U ta h . P ap ers, P eab o d v M u seu m of
A m erican A rchaeologv a n d E thnology, H a rv a rd 21.
E h ric h , R o b e rx W.
1950 Som e reflections on archeological in te rp re ta tio n . A m erican A nthropologist 5 2 :4 6 8 -
82.
F o r d , J a m e s A.
1952 M e asu rem en ts of som e prehistoric design d ev elo p m en ts in th e so u th e a ste rn sta te s.
A nth ro p o lo g ical P a p e rs. A m erican M useum of N a tu r a l H isto ry 44, P t. 3.
K l ü c k h o h n , C lyde
1939 T h e place of th e o ry in anthropological studies. P h ilo so p h y of Science 6, N o . 3:
328-44.
K r ie g e r , A l e x D .
1944 T h e typological co n cep t. A m erican A n tiq u ity 9, N o. 3 :2 7 1 -8 8 .
K r o e b e r , A . I .. a n d C l y d e K l u c k iio h n
1952 C u ltu re , a critic al review of concepts an d definitions. P a p e rs, P eabodv M u seu m oí
A m erican A rchaeology an d E thnology, H a rv a rd 47, N o. 1.
L othrop, S. K .
1926 P o tte ry of C o s ta R ic a an d N icaragua. Vol. I. C o n trib u tio n s, M useum of th e A m eri­
can In d ia n , H ey e F o u n d a tio n , N ew Y ork.
M 'l l k e , W i l h e l m
1949 T h e q u a n tita tiv e d istrib u tio n of cu ltu ral sim ilarities a n d th eir carto g ra p h ic repre-
se n ta tio n . A m erican A nthropologist 51:237-52.
P h il l ip s , P h il ip , J am es A. F ord a n d J ames B . G r if f in
1951 A rchaeological su rv ey in th e lower M ississippi alluvial valley, 1940-1947. P ap ers,
P e a b o d y M u seu m of A m erican A rchaeology an d E th n o lo g y , H a rv a rd 25.
R o u s e , I r v in g
1939 P re h isto ry in H a iti, a stu d y in m ethod. Y ale U n iv ersity P u b licatio n s in A n th ro -
pology, N o . 21.
 A merican A nthropologist [56, 1954
S pa ttld in g , A l b e r t C.
1953 S ta tistic a l tech n iq u es for th e discovery of a rtifa c t types. A m erican A n tiq u ity 18:
305-13.
T aylor, W a l te r W .
1948 A s tu d y of archeology. A m erican A nthropologist 50, N o. 3, P t. 2 (“ A m erican An-
thropological A ssociation M em oirs,” N o. 69).
W elley , G o rd o n
1953 P re h isto ric se ttle m e n t p a tte rn s in th e V irú Valley, P erú. B ureau of A m erican F.th-
nology, B ull. 155.

Types of Types
These com m ents are intended to extend F ord’s excellent clarification of
“ ty p e ” by distinguishing several meanings of the term and showing th a t each
has special significance relative to problem. I am concerned here with four
meanings of “ ty p e ”— there are, of coursej m any other meanings— which I
shall desígnale as “ morphological,” “ historical-index,” “ functional,” and
“ cultural.” T he same c ultural phenomenon m ay be classified as any or all oí
these types. F ord illustrates his points with the imaginary people of the
island of G am m a-gam m a. Although the illustration is ethnographic, the
heuristic meaning of his types applies more often to archeological than to
ethnographic problems; for F o r d ’s types are those I desígnate “ morphological”
an d “historical-index,” being used primarily to determine the time-space oc-
currence of cultural phenomena.
The “ morphological” type is the most elementary kind, since it is based
solely on form— on physical or external properties. When the use or cultural
significance of an object or practice is unknown, a descriptive label is necessary.
Thus, “ stone balls” which characterize the Northern Periphery of the Anasazi
area is a taxonomic label, which, to quote F o rd ’s remark to me, represents the
lowest order of archeological procedure. W h at were proved later to be ball
courts in Arizona were a t first cl^ssed under such noncom mittal headings as
“ large, basin-like depressions,” which, for all th a t anvone knew, m ight have
been dance plazas or reservoirs.
T he second, or “ historical-index” type, is defined by form, but, whereas
the morphological type .is considered as a characteristic of the culture, this
second type has chronological, not cultural, significance. I t is a time-marker.
P o tte ry is an outsta n d in g example of this second type. Although p o tte ry is of
course a p art of thé cultural inventory, its various elements such as clay,
shape, design, and the like which combine to define ceramic types are widely
used to distinguish chronological an d aerial differences. T h a t is, changes in
ceraríiic styles, like changes in soil types, pollen count, frequencies of key
flora and fauna, and otuer means of dating, have noncultural significance.
M a n y other cultural features m ay be used as historical Índices. The líthic com-
plexes of the Folsom, Y am an, Silver Lake and other early periods include
projectile points which strongiy indícate th a t hunting with spears was an
essential feature of the cultures. B u t while the forms of the projectile points in