THE RELATION OF HABITUALTHOUGHT AND BEHAVIORTO LANGUAGE

BENJAMIN LEE WHORF

[EDITOR'S FOREWORD : Few people have been as well qualified as Benjamin Lee Whorf to ex-plain, from personal knowledge and study, what is meant by the expression, 'the structure oflanguage .' An outstanding authority on the Mayan and Aztec civilizations and on American Indianlanguages, be was intimately acquainted with languages whose basic structures were totally unlikethose of the Indo-European languages. What is even more to the point, Mr . Whorf was extraor-dinarily sensitive to the non-linguistic consequences of linguistic behavior. It is obvious from hiswritings that, all the time he was investigating languages, whether among the Pueblo villages ofCentral Mexico or among the Hopi in Arizona, he must have been watching what was going on-what actions, what attitudes, what events accompanied or resulted from the utterances he wasso carefully recording .Among Mr. Whorf's many contributions to language study, the highest place must be accorded

-at least in the eyes of those interested in general semantics-to his demonstration throughcomparative linguistics that our day-to-day orientations in life, to say nothing of our 'reasoningprocesses' and our 'philosophies,' rest upon the structure of the language which we happen tohave inherited. Our 'common sense,' our most basic 'intuitions' into the 'nature of things,' ourdichotomy of 'form' and 'substance,' our notions of 'time,' 'space,' and 'matter,' and even ourlife-habits and our social institutions are shaped to a degree hitherto unsuspected, Mr. Whorfbelieved, by the structuralizations which our languages impose upon the flux of experience.

'The Relation of Habitual Thought and Behavior to Language' combines Mr . Whorf's experi-ences as anthropologist, linguist, grammarian, and fire insurance executive . Born in 1897 inWinthrop, Mass ., he was a graduate of M.I .T., and served as a private in the engineering corpsduring World War I . In 1919 he joined the Hartford Fire Insurance Company, and was assistantsecretary of the company at the time of his death, July 26, 1941 . He began his study of Aztecand Mayan cultures, as a hobby, in 1925 . Within a few years he had become one of the nation'sleading Americanists . Many of his articles, the results of field work in Mexico and the Southwestas well as of private study, were published in Technology Review . Two of his articles, 'Lan-guages and Logic,' published in Papers from the Second American Congress of General Se-mantics, and 'Science and Linguistics,' reprinted as an appendix to Hayakawds Language inAction, are already familiar to students of general semantics . A more complete biography withbibliography will be found in the National Cyclopedia of American Biography.

The present article is reprinted by permission from Language, Culture, and Personality: Essaysin Memory of Edward Sapir (Menasha, Wisconsin, 1941) .]

THERE will probably be general assentto the, proposition that an accepted

pattern of using words is often prior tocertain lines of thinking and forms ofbehavior, but he who assents often seesin such a statement nothing more than aplatitudinous recognition of the hypnoticpower of philosophical and learned term-inology on the one hand or of catchwords,

19 7

slogans, and rallying-cries on the other.To see only thus far is to miss the pointof one of the important interconnectionswhich Sapir saw between language, cul-ture, and psychology, and succinctly ex-pressed in the introductory quotation .* Itis not so much in these special uses oflanguage as in its constant ways of arrang-

* See next page.

ETC. : A REVIEW OF GENERAL SEMANTICS

ing data and its most ordinary every-dayanalysis of phenomena that we need torecognize the influence it has on otheractivities, cultural and personal .

The Name o f the Situationas Affecting Behavior

I came in touch with an aspect of thisproblem before I had studied under Dr .Sapir, and in a field usually consideredremote from linguistics . It was in thecourse of my professional work for a fireinsurance company, in which I undertookthe task of analyzing many hundreds ofreports of circumstances surrounding thestart of fires, and in some cases, of ex-plosions. My analysis was directed towardpurely physical conditions, such as defec-tive wiring, presence or lack of air spacesbetween metal flues and woodwork, etc .,and the results were presented in theseterms. Indeed it was undertaken with nothought that any other significances wouldor could be revealed . But in due course itbecame evident that not only a physicalsituation qua physics, but the meaning ofthat situation to people, was sometimes afactor, through the behavior of the peo-ple, in the start of the fire . And thisfactor of meaning was clearest when itwas a linguistic meaning, residing in thename or the linguistic description com-monly applied to the situation . Thusaround a storage of what are called 'gaso-line drums' behavior will tend to a certaintype, that is, great care will be exercised ;while around a storage of what are called'empty gasoline drums' it will tend to bedifferent-careless, with little repressionof smoking or of tossing cigarette stubsabout. Yet the 'empty' drums are perhapsthe more dangerous, since they containexplosive vapor . Physically the situation ishazardous, but the linguistic analysis ac-cording to regular analogy must employ

198

the word 'empty,' which inevitably sug-gests lack of hazard. The word 'empty' isused in two linguistic patterns : (1) as avirtual synonym for 'null and void, nega-tive, inert,' (2) applied in analysis ofphysical situations without regard to, e .g.,vapor, liquid vestiges, or stray rubbish, inthe container . The situation is named inone pattern (2) and the name is then'acted out' or 'lived up to' in another (1) ;this being a general formula for the lin-guistic conditioning of behavior into haz-ardous forms .

In a wood distillation plant the metalstills were insulated with a composition

'Human beings do not live in the ob-jective world alone, nor alone in theworld o f social activity as ordinarily un-derstood, but are very much at the mercyo f the particular language which has be-come the medium o f expression for theirsociety. It is quite an illusion to imaginethat one adjusts to reality essentially with-out the use o f language and that languageis merely an incidental means of solvingspecific problems o f communication or re-flection. The fact o f the matter is that the'real world' is to a large extent uncon-sciously built up on the language habits o f

the group . . . . We see and hear and other-wise experience very largely as we do be-cause the language habits of our commu-nity predispose certain choices o f interpre-tation :-EnwARw SAPIR, 'The Status ofLinguistics as a Science,' Language, Vol .V, pp. 209-210 (1929) .

prepared from limestone and called atthe plant 'spun limestone .' No attemptwas made to protect this covering fromexcessive heat or the contact of flame.After a period of use the fire below oneof the stills spread to the 'limestone,'

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE

which to everyone's great surprise burnedvigorously. Exposure to acetic acid fumesfrom the stills had converted part of thelimestone (calcium carbonate) to calciumacetate . This when heated in a fire de-composes, forming inflammable acetone .Behavior that tolerated fire close to thecovering was induced by use of the name'limestone,' which because it ends in'-stone' implies noncombustibility.

A huge iron kettle of boiling varnishwas observed to be overheated, nearingthe temperature at which it would ignite.The operator moved it off the fire andran it on its wheels to a distance, butdid not cover it . In a minute or so thevarnish ignited . Here the linguistic influ-ence is more complex ; it is due to themetaphorical objectifying (of which morelater) of `cause' as contact or the spatialjuxtaposition of 'things'-to analyzing thesituation as 'on' versus 'off' the fire . Inreality the stage when the external firewas the main factor had passed ; the over-heating was now an internal process ofconvection in the varnish from the in-tensely heated kettle, and still continuedwhen 'off' the fire.

An electric glow heater on the wallwas little used, and for one workman hadthe meaning of a convenient coat-hanger .At night a watchman entered and snappeda switch, which action he verbalized as'turning on the light.' No light appeared,and this result he verbalized as 'light isburned out .' He could not see the glowof the heater because of the old coat hungon it. Soon the heater ignited the coat,which set fire to the building.

A tannery discharged waste water con-taining animal matter into an outdoorsettling basin partly roofed with woodand partly open . This situation is one thatordinarily would be verbalized as 'poolof water.' A workman had occasion tolight a blow-torch nearby, and threw his

199

match into the water . But the decompos-ing waste matter was evolving gas underthe wood cover, so that the setup wasthe reverse of 'watery.' An instant flareof flame ignited the woodwork, and thefire quickly spread into 'the adjoiningbuilding.

A drying room for hides was arrangedwith a blower at one end to make a cur-rent of air along the room and thenceoutdoors through a vent at the other end .Fire started at a hot bearing on theblower, which blew the flames directlyinto the hides and fanned them along theroom, destroying the entire stock . Thishazardous setup followed naturally fromthe term 'blower' with its linguistic equiva-lence to 'that which blows,' implying thatits function necessarily is to 'blow.' Alsoits function is verbalized as 'blowing airfor drying,' overlooking that it can blowother things, e.g ., flames and sparks. Inreality a blower simply makes a currentof air and can exhaust as well as blow.It should have been installed at the ventend to draw the air over the hides, thenthrough the hazard (its own casing andbearings) and thence outdoors .

Beside a coal-fired melting pot for leadreclaiming was dumped a pile of 'scraplead'-a misleading verbalization, for itconsisted of the lead sheets of old radiocondensers, which still had paraffin paperbetween them. Soon the paraffin blazedup and fired the roof, half of which wasburned off.

Such examples, which could be greatlymultiplied, will suffice to show how thecue to a certain line of behavior is oftengiven by the analogies of the linguisticformula in which the situation is spokenof, and by which to some degree it isanalyzed, classified, and allotted its placein that world which is 'to a large extentunconsciously built up on the languagehabits of the group .' And we always as-

ETC. : A REVIEW OF GENERAL SEMANTICS

sume that the linguistic analysis made byour group reflects reality better than itdoes .

Grammatical Patterns as Inter-pretations o f Experience

The linguistic material in the aboveexamples is limited to single words,phrases and patterns of limited range .One cannot study the behavioral com-pulsiveness of such material without sus-pecting a much more far-reaching com-pulsion from large-scale patterning ofgrammatical categories, such as plurality,gender and similar classifications (ani-mate, inanimate, etc.), tenses, voices, andother verb forms, classifications of thetype of 'parts of speech,' and the matterof whether a given experience is denotedby a unit morpheme, an inflected word,or a syntactical combination . A categorysuch as number (singular vs . plural) isan attempted interpretation of a wholelarge order of experience, virtually of theworld or of nature ; it attempts to say howexperience is to be segmented, what ex-perience is to be called 'one' and what'several .' But the difficulty of appraisingsuch a far-reaching influence is great be-cause of its background character, becauseof the difficulty of standing aside fromour own language, which is a habit and acultural non est disputandum, and scru-tinizing it objectively. And if we take avery dissimilar language, this languagebecomes a part of nature, and we evendo to it what we have already done tonature. We tend to think in our ownlanguage in order to examine the exoticlanguage. Or we find the task of un-raveling the purely morphological intric-acies so gigantic that it seems to absorball else. Yet the problem, though difficult,is feasible ; and the best approach isthrough an exotic language, for in itsstudy we are at long last pushed willy-

200

filly out of our ruts. Then we find thatthe exotic language is a mirror held upto our own .

In my study of the Hopi language,what I now see as an opportunity to workon this problem was first thrust upon mebefore I was clearly aware of the prob-lem. The seemingly endless task of de-scribing the morphology did finally end .Yet it was evident, especially in the lightof Sapir's lectures on Navaho, that thedescription of the language was far fromcomplete. I knew for example the mor-phological formation of plurals, but nothow to use plurals. It was evident that thecategory of plural in Hopi was not thesame thing as in English, French, orGerman. Certain things that were pluralin these languages were singular in Hopi .The phase of investigation which nowbegan consumed nearly two more years.

The work began to assume the characterof a comparison between Hopi and west-ern European languages . It also becameevident that even the grammar of Hopibore a relation to Hopi culture, and thegrammar of European tongues to ourown 'western' or 'European' culture . Andit appeared that the interrelation broughtin those large subsummations of experi-ence by language, such as our terms'time,' 'space,' 'substance,' and 'matter .'Since with respect to the traits comparedthere is little difference between English,French, German, or other European lan-guages with the possible (but doubtful)exception of Balto-Slavic and non-Indo-European, I have lumped these languagesinto one group called SAE, or 'StandardAverage European .'

That portion of the whole investigationhere to be reported may be summed up intwo questions : (1) Are our own conceptsof 'time,' 'space,' and 'matter' given insubstantially the same form by experienceto all men, or are they in part conditioned

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE

by the structure of particular languages?(2) Are there traceable affinities between(a) cultural and behavioral norms and(b) large-scale linguistic patterns? Ishould be the last to pretend that thereis anything so definite as 'a correlation'between culture and language, and espe-cially between ethnological rubrics suchas 'agricultural,' 'hunting,' etc ., and lin-guistic ones like 'inflected,' 'synthetic,' or'isolating." When I began the study theproblem was by no means so clearly form-ulated and I had little notion that the an-swers would turn out as they did .

Plurality and Numerationin SAE and Hopi

In our language, that is SAE, pluralityand cardinal numbers are applied in twoways : to real plurals and imaginaryplurals . Or more exactly if less tersely :perceptible spatial aggregates and meta-phorical aggregates . We say 'ten men'and also 'ten days.' Ten men either areor could be objectively perceived as ten,ten in one group-perception-ten menon a street corner, for instance . But 'tendays' cannot be objectively experienced .We experience only one day, to-day ; theother nine (or even all ten) are some-thing conjured up from memory orimagination. If 'ten days' be regarded asa group it must be as an 'imaginary,' men-tally constructed group . Whence comesthis mental pattern? Just as in the case of

' We have plenty of evidence that this is notthe case. Consider only the Hopi and the Ute,with languages that on the overt morphologicaland lexical level are as similar as, say, Englishand German. The idea of 'correlation' betweenlanguage and culture, in the generally acceptedsense of correlation, is certainly a mistaken one .

' As we say, 'ten at the same time,' showingthat in our language and thought we restate thefact of group-perception in terms of a concept'time,' the large linguistic component of whichwill appear in the course of this paper .

2 0 1

the fire-causing errors, from the fact thatour language confuses the two differentsituations, has but one pattern for both.When we speak of ten steps forward, tenstrokes on a bell, or any similarly de-scribed cyclic sequence, 'times' of any sort,we are doing the same thing as with 'days .'Cyclicity brings the response of imaginaryplurals. But a likeness of cyclicity to ag-gregates is not unmistakably given by ex-perience prior to language, or it would befound in all languages, and it is not .Our awareness of, time and cyclicity

does contain something immediate andsubjective-the basic sense of 'becom-ing later and later.' But in the habitualthought of us SAE people this is coveredunder something quite different, whichthough mental should not be called sub-jective. I call it objectified, or imaginary,because it is patterned on the outer world .It is this that reflects our linguistic usage .Our tongue makes no distinction betweennumbers counted on discrete entities andnumbers that are simply counting itself.Habitual thought then assumes that in thelatter case the numbers are just as muchcounted on something as in the former .This is objectification. Concepts of timelose contact with the subjective experienceof 'becoming later' and are objectified ascounted quantities, especially as lengths,made up of units as a length can be visiblemarked off into inches . A 'length of time'is envisioned as a row of similar units,like a row of bottles .

In Hopi there is a different linguisticsituation. Plurals and cardinals are usedonly for entities that form or can forman objective group. There are no imag-inary plurals, but instead ordinals usedwith singulars. Such an expression as 'tendays' is not used. The equivalent statementis an operational one that reaches one dayby a suitable count. 'They stayed ten days'becomes 'they stayed until the eleventh

ETC. : A REVIEW OF GENERAL SEMANTICS

day' or 'they left after the tenth day .''Ten days is greater than nine days' be-comes 'the tenth day is later than theninth.' Our 'length of time' is not re-garded as a length but as a relation be-tween two events in lateness. Instead ofour linguistically promoted objectificationof that datum of consciousness we call'time,' the Hopi language has not laiddown any pattern that would cloak thesubjective 'becoming later' that is the es-sence of time.

Nouns of Physical Quantityin SAE and Hopi

We have two kinds of nouns denotingphysical things ; individual nouns, andmass nouns, e.g ., water, milk, wood, gran-ite, sand, flour, meat . Individual nouns de-note bodies with definite outlines: a tree,a stick, a man, a hill. Mass nouns denotehomogeneous continua without impliedboundaries. The distinction is markedby linguistic form ; e.g., mass nouns lackplurals,s in English drop articles, and inFrench take the partitive article du, de la,des . The distinction is more widespreadin language than in the observable appear-ance of things . Rather few natural occur-rences present themselves as unboundedextents ; air of course, and often water,rain, snow, sand, rock, dirt, grass. Wedo not encounter butter, meat, cloth, iron,glass, or most 'materials' in such kind ofmanifestation, but in bodies small or largewith definite outlines. The distinction is

* It is no exception to this rule of lacking aplural that a mass noun may sometimes coin-cide in lexeme with an individual noun that ofcourse has a plural ; e .g., 'stone' (no p1 .) with'a stone' (pl. 'stones') . The plural form denot-ing varieties, e .g., 'wines' is of course a differ-ent sort of thing from the true plural ; it is acurious outgrowth from the SAE mass nouns,leading to still another sort of imaginary aggre-gates, which will have to be omitted from thispaper .

2 0 2

somewhat forced upon our description ofevents by an unavoidable pattern in lan-guage. It is so inconvenient in a greatmany cases that we need some way of in-dividualizing the mass noun by furtherlinguistic devices. This is partly done bynames of body-types : stick of wood, pieceof cloth, pane of glass, cake of soap ; also,and even more, by introducing names ofcontainers though their contents be thereal issue : glass of water, cup of coffee,dish of food, bag of flour, bottle of beer .These very common container-formulas,in which 'of' has an obvious, visually per-ceptible meaning ('contents'), influenceour feeling about the less obvious type-body formulas : stick of wood, lump ofdough, etc. The formulas are very similar :individual noun plus a similar relator(English 'of') . In the obvious case thisrelator denotes contents. In the inobviousone it suggests contents. Hence the lumps,chunks, blocks, pieces, etc ., seem to con-tain something, a 'stuff,' 'substance,' or'mattei that answers to the water, coffee,or flour in the container formulas. So withSAE people the philosophic 'substance'and 'matter' are also the naive idea ; theyare instantly acceptable, 'common sense .'It is so through linguistic habit. Our lan-guage patterns often require us to namea physical thing by a binomial that splitsthe reference into a formless item plus aform.

Hopi is again different. It has a form-ally distinguished class of nouns . But thisclass contains no formal sub-class of massnouns. All nouns have an individual senseand both singular and plural forms .Nouns translating most nearly our massnouns still refer to vague bodies orvaguely bounded extents. They imply in-definiteness, but not lack, of outline andsize. In specific statements 'water' meansone certain mass or quantity of water, notwhat we call 'the substance water.' Gen-

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE

erality of statement is conveyed throughthe verb or predicator, not the noun .Since nouns are individual already theyare not individualized either by type-bodies or names of containers, if there isno special need to emphasize shape orcontainer. The noun itself implies a suit-able type-body or container. One says, not'a glass of water' but ke.yi 'a water,' not'a pool of water' but pa.he, 4 not 'a dishof corn-flour' but ngemni 'a (quantity of)corn-flour,' not 'a piece of meat' but sik"i'a meat.' The language has neither needfor nor analogies on which to build theconcept of existence as a duality of form-less item and form. It deals with formless-ness through other symbols than nouns .

Phases of Cycles in SAE and HopiSuch terms as summer, winter, Sep-

tember, morning, noon, sunset, are withus nouns, and have little formal linguisticdifference from other nouns . They can besubjects or objects, and we say 'at' sun-set or 'in' winter just as we say at a corneror in an orchard.5 They are pluralized andnumerated like nouns of physical objects,as we have seen. Our thought about thereferents of such words hence becomesobjectified . Without objectification itwould be a subjective experience of realtime, i .e. of the consciousness of 'becom-ing later and later'-simply a cyclic phasesimilar to an earlier phase in that ever-later-becoming duration . Only by imag-

' Hopi has two words for water-quantities ;ke.yi and pa-he. The difference is somethinglike that between 'stone' and 'rock' in English,pa-he implying greater size and 'wildness' ;flowing water, whether or not out-doors or innature, is pa-be, so is 'moisture .' But unlike'stone' and 'rock,' the difference is essential,not pertaining to a connotative margin, and thetwo can hardly ever be interchanged .

'To be sure there are a few minor differencesfrom other nouns, in English for instance inthe use of the articles .

2 03

ination can such a cyclic phase be set be-side another and another in the mannerof a spatial (i .e . visually perceived) con-figuration. But such is the power of lin-guistic analogy that we do so objectifycyclic phasing . We do it even by saying'a phase' and 'phases' instead of, e.g.,'phasing.' And the pattern of individualand mass nouns, with the resulting bi-nomial formula of formless item plusform, is so general that it is implicit forall nouns, and hence our very generalizedformless items like 'substance,' 'matter,'by which we can fill out the binomialfor an enormously wide range of nouns .But even these are not quite generalizedenough to take in our phase nouns. Sofor the phase nouns we have made aformless item, 'time.' We have made itby using 'a time,' i .e. an occasion or aphase, in the pattern of a mass noun, justas from 'a summer' we make 'summer' inthe pattern of a mass noun . Thus with ourbinomial formula we can say and think'a moment of time,' 'a second of time,''a year of time.' Let me again point outthat the pattern is simply that of 'a bottleof milk' or 'a piece of cheese.' Thus weare assisted to imagine that 'a summer'actually contains or consists of such-and-such a quantity of 'time .'

In Hopi however all phase terms, likesummer, morning, etc ., are not nouns buta kind of adverb, to use the nearest SAEanalogy. They are a formal part of speechby themselves, distinct from nouns, verbs,and even other Hopi 'adverbs.' Such aword is not a case form or a locative pat-tern, like 'des Abends' or 'in the morn-ing.' It contains no morpheme like oneof 'in the house' or 'at the tree.'e It means

'Year' and certain combinations of 'year'with name of season, rarely season names alone,can occur with a locative morpheme 'at,' butthis is exceptional . It appears like historicaldetritus of an earlier different patterning, orthe effect of English analogy, or both.

ETC. : A REVIEW OF GENERAL SEMANTICS

'when it is morning or 'while morning-phase is occurring.' These 'temporals' arenot used as subjects or objects, or at alllike nouns. One does not say 'it's a hotsummer' or 'summer is hot ;' summer isnot hot, summer is only when conditionsare hot, when heat occurs. One does notsay 'this summer,' but 'summer now' or'summer recently.' There is no objec-tification, as a region, an extent, a quan-tity, of the subjective duration-feeling.Nothing is suggested about time exceptthe perpetual 'getting later' of it. And sothere is no basis here for a formless itemanswering to our 'time .'

Temporal Forms o f Verbsin SAE and Hopi

The three-tense system of SAE verbscolors all our thinking about time . Thissystem is amalgamated with that largerscheme of objectification of the subjectiveexperience of duration already noted inother patterns-in the binomial formulaapplicable to nouns in general, in tem-poral nouns, in plurality and numeration .This objectification enables us in imag-ination to 'stand time units in a row .'Imagination of time as like a row har-monizes with a system of three tenses ;whereas a system of two, an earlier and alater, would seem to correspond better tothe feeling of duration as it is experi-enced. For if we inspect consciousness wefind no past, present, future, but a unityembracing complexity . Everything is inconsciousness, and everything in con-sciousness is, and is together. There is init a sensuous and a non-sensuous . We maycall the sensuous-what we are seeing,hearing, touching-the 'present' while inthe non-sensuous the vast image-world ofmemory is being labelled 'the past' andanother realm of belief, intuition, and un-certainty 'the future ;' yet sensation, mem-

2 04

ory, foresight, all are in consciousness to-gether-one is not 'yet to be' nor another'once but no more.' Where real timecomes in is that all this in consciousnessis 'getting later,' changing certain relationsin an irreversible manner . In this 'latering'or 'durating' there seems to me to be aparamount contrast between the newest,latest instant at the focus of attention andthe rest-the earlier. Languages by thescore get along well with two tense-likeforms answering to this paramount rela-tion of later to earlier. We can of courseconstruct and contemplate in thought asystem of past, present, future, in the ob-jectified configuration of points on a line .This is what our general objectificationtendency leads us to do and our tensesystem confirms .

In English the present tense seems theone least in harmony with the paramounttemporal relation . It is as if pressed intovarious and not wholly congruous duties .One duty is to stand as objectified middleterm between objectified past and objecti-fied future, in narration, discussion, argu-ment, logic, philosophy . Another is to de-note inclusion in the sensuous field : 'Isee him.' Another is for nomic, i .e. cus-tomarily or generally valid, statements :'We see with our eyes .' These varied usesintroduce confusions of thought, of whichfor the most part we are unaware .

Hopi, as we might expect, is differenthere too. Verbs have no ' tenses' like ours,but have validity-forms ('assertions'), as-pects, and clause-linkage forms (modes),that yield even greater precision of speech.The validity-forms denote that the speaker(not the subject) reports the situation(answering to our past and present) orthat he expects it (answering to our fu-ture) 7 or that he makes a nomic statement

° The expective and reportive assertions con-trast according to the 'paramount relation .' Theexpective expresses anticipation existing earlier

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE

(answering to our nomic present) . Theaspects denote different degrees of dura-tion and different kinds of tendency 'dur-ing duration.' As yet we have notednothing to indicate whether an event issooner or later than another when bothare reported . But need for this does notarise until we have two verbs, i .e. twoclauses . In that case the 'modes' denoterelations between the clauses, includingrelations of later to earlier and of simul-taneity . Then there are many detachedwords that express similar relations, sup-plementing the modes and aspects. Theduties of our three-tense system and itstripartite linear objectified 'time' are dis-tributed among various verb categories,all different from our tenses ; and there isno more basis for an objectified time inHopi verbs than in other Hopi patterns ;although this does not in the least hinderthe verb forms and other patterns frombeing closely adjusted to the pertinentrealities of actual situations .

Duration, Intensity, and Tendencyin SAE and Hopi

To fit discourse to manifold actual situ-ations all languages need to express dura-tions, intensities, and tendencies . It ischaracteristic of SAE and perhaps of manyother language-types to express themmetaphorically. The metaphors are thoseof spatial extension, i.e. of size, number(plurality), position, shape, and motion .

than objective fact, and coinciding with objec-tive fact later than the status quo of thespeaker, this status quo, including all the sub-summation of the past therein, being expressedby the reportive . Our notion 'future' seems torepresent at once the earlier (anticipation) andthe later (afterwards, what will be), as Hopishows. This paradox may hint of how elusivethe mystery of real time is, and how artificiallyit is expressed by a linear relation of past-present-future .

2 0 5

We express duration by long, short, great,much, quick, slow, etc . ; intensity bylarge, great, much, heavy, light, high, low,sharp, faint, etc. ; tendency by more, in-crease, grow, turn, get, approach, go,come, rise, fall, stop, smooth, even, rapid,slow, and so on through an almost in-exhaustible list of metaphors that wehardly recognize as such since they arevirtually the only linguistic media avail-able. The non-metaphorical terms in thisfield, like early, late, soon, lasting, intense,very, tending, are a mere handful, quiteinadequate to the needs .

It is clear how this condition 'fits in .'It is part of our whole scheme of objecti-fying-imaginatively spatializing qualitiesand potentials that are quite non-spatial(so far as any spatially-perceptive sensescan tell us) . Noun-meaning (with us)proceeds from physical bodies to referentsof far other sort. Since physical bodiesand their outlines in perceived space aredenoted by size and shape terms andreckoned by cardinal numbers and plurals,these patterns of denotation and reckon-ing extend to the symbols of non-spatialmeanings, and so suggest an imaginaryspace. Physical shapes move, stop, rise,sink, approach, etc., in perceived space ;why not these other referents in theirimaginary space? This has gone so far thatwe can hardly refer to the simplest non-spatial situation without constant resort tophysical metaphors . I 'grasp' the 'thread'of another's arguments, but if its 'level' is'over my head' my attention may 'wander'and 'lose touch' with the 'drift' of it, sothat when he 'comes' to his 'point' wediffer 'widely,' our 'views' being indeed so'far apart' that the 'things' he says 'ap-pear' 'much' too arbitrary, or even 'a lot'of nonsense!

The absence of such metaphor fromHopi speech is striking . Use of spaceterms when there is no space involved is

ETC. : A REVIEW OF GENERAL SEMANTICS

not there-as if on it had been laid thetaboo teetotal! The reason is clear whenwe know that Hopi has abundant conju-gational and lexical means of expressingduration, intensity, and tendency directlyas such, and that major grammatical pat-terns do not, as with us, provide analogiesfor an imaginary space. The many verb'aspects' express duration and tendency ofmanifestations, while some of the 'voices'express intensity, tendency, and durationof causes or forces producing manifesta-tions. Then a special part of speech, the'tensors,' a huge class of words, denotesonly intensity, tendency, duration, andsequence. The function of the tensors isto express intensities, 'strengths,' and howthey continue or vary, their rate-of-change ;so that the broad concept of intensity,when considered as necessarily alwaysvarying and/or continuing, includes alsotendency and duration. Tensors conveydistinctions of degree, rate, constancy,repetition, increase and decrease of in-tensity, immediate sequence, interruptionor sequence after an interval, etc ., alsoqualities of strengths, such as we shouldexpress metaphorically as smooth, even,hard, rough. A striking feature is theirlack of resemblance to the terms of realspace and movement that to us 'meanthe same.' There is not even more thana trace of apparent derivation from spaceterms.8 So while Hopi in its nouns seemshighly concrete, here in the tensors it be-comes abstract almost beyond our ownpower to follow.

'One such trace is that the tensor 'long induration,' while quite different from the adjec-tive 'long' of space, seems to contain the sameroot as the adjective 'large' of space . Anotheris that 'somewhere' of space used with certaintensors means 'at some indefinite time .' Pos-sibly however this is not the case and it is onlythe tensor that gives the time element, so that'somewhere' still refers to space and that underthese conditions indefinite space means simply

206

Habitual Thought in SAEand Hopi

The comparison now to be made be-tween the habitual thought worlds of SAEand Hopi speakers is of course incom-plete. It is possible only to touch uponcertain dominant contrasts that appear tostem from the linguistic differences al-ready noted . By 'habitual thought' and'thought world' I mean more than simplylanguage, i .e ., than the linguistic patternsthemselves. I include all the analogicaland suggestive value of the patterns (e.g.,our 'imaginary space' and its distant im-plications), and all the give-and-take be-tween language and the culture as awhole, wherein is a vast amount that is notlinguistic yet shows the shaping influenceof language . In brief, this 'thought world'is the microcosm that each man carriesabout within himself, by which he meas-ures and understands what he can of themacrocosm .

The SAE microcosm has analyzed re-ality largely in terms of what it calls'things' (bodies and quasi-bodies) plusmodes of extensional but formless exist-ence that it calls 'substance' or 'matter .'It tends to see existence through a bi-nomial formula that expresses any existentas a spatial form plus a spatial formlesscontinuum related to the form as contentis related to the outlines of its container .Non-spatial existents are imaginativelyspatialized and charged with similar im-plications of form and continuum.The Hopi microcosm seems to have

analyzed reality largely in terms of events

general applicability regardless of either timeor space. Another trace is that in the temporal(cycle word) 'afternoon' the element meaning'aftei is derived from the verb 'to separate .'There are other such traces, but they are fewand exceptional, and obviously not like our ownspatial metaphorizing.

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE

(or better 'eventing'), referred to in twoways, objective and subjective. Objec-tively, and only if perceptible physicalexperience, events are expressed mainlyas outlines, colors, movements, and otherperceptive reports . Subjectively, for boththe physical and non-physical, events areconsidered the expression of invisible in-tensity-factors, on which depend their sta-bility and persistence, or their fugitivenessand proclivities. It implies that existentsdo not 'become later and later' all in thesame way ; but some do so by growing,like plants, some by diffusing and vanish-ing, some by a procession of metamor-phoses, some by enduring in one shapetill affected by violent forces . In the na-ture of each existent able to manifest asa definite whole is the power of its ownmode of duration ; its growth, decline,stability, cyclicity, or creativeness . Every-thing is thus already 'prepared' for theway it now manifests by earlier phases,and what it will be later, partly has been,and partly is in act of being so 'prepared .'An emphasis and importance rests on thispreparing or being prepared aspect of theworld that may to the Hopi correspond tothat 'quality of reality' that 'matter' or'stuff' has for us .

Habitual Behavior Featureso f Hopi Culture

Our behavior, and that of Hopi, can beseen to be coordinated in many ways tothe linguistically-conditioned microcosm.As in my fire case-book, people act aboutsituations in ways which are like the waysthey talk about them . A characteristic ofHopi behavior is the emphasis on prepara-tion. This includes announcing and get-ting ready for events well beforehand,elaborate precautions to insure persistenceof desired conditions, and stress on goodwill as the preparer of right results . Con-sider the analogies of the day-counting

207

pattern alone. Time is mainly reckoned'by day' (talk, -tala) or 'by night' (tok),which words are not nouns but tensors,the first formed on a root 'light, day,' thesecond on a root 'sleep .' The count is byordinals. This is not the pattern of count-ing a number of different men or things,even though they appear successively, foreven then they could gather into an assem-blage. It is the pattern of counting succes-sive reappearances of the same man orthing, incapable of forming an assemblage.The analogy is not to behave about day-cyclicity as to several men ('several days'),which is what we tend to do, but to be-have as to the successive visits of the sameman. One does not alter several men byworking upon just one, but one can pre-pare and so alter the later visits of thesame man by working to affect the visithe is making now. This is the way theHopi deal with the future-by workingwithin a present situation which is ex-pected to carry impresses, both obviousand occult, forward into the future eventof interest. One might say that Hopi so-ciety understands our proverb 'Well be-gun is half done,' but not our To-mor-row is another day.' This may explainmuch in Hopi character .

This Hopi preparing behavior may beroughly divided into announcing, outerpreparing, inner preparing, covert par-ticipation, and persistence. Announcing,or preparative publicity, is an importantfunction in the hands of a special official,the Crier Chief. Outer preparing is prepa-ration involving much visible activity, notall necessarily directly useful within ourunderstanding. It includes ordinary prac-tising, rehearsing, getting ready, introduc-tory formalities, preparing of special food,etc. (all of these to a degree that mayseem over-elaborate to us), intensive sus-tained muscular activity like running, rac-ing, dancing, which is thought to increase

ETC . : A REVIEW OF GENERAL SEMANTICS

the intensity of development of events(such as growth of crops), mimetic andother magic, preparations based on eso-teric theory involving perhaps occult in-struments like prayer sticks, prayer feath-ers, and prayer meal, and finally the greatcyclic ceremonies and dances, which havethe significance of preparing rain andcrops. From one of the verbs meaning'prepare' is derived the noun for 'harvest'or 'crop :' na'twani 'the prepared' or the'in preparation.' 9

Inner preparing is use of prayer andmeditation, and at lesser intensity goodwishes and good will, to further desiredresults. Hopi attitudes stress the power ofdesire and thought. With their 'microcosm'it is utterly natural that they should. De-sire and thought are the earliest, andtherefore the most important, most criti-cal and crucial, stage of preparing . More-over, to the Hopi, one's desires andthoughts influence not only his own ac-tions, but all nature. This too is whollynatural. Consciousness itself is aware ofwork, of the feel of effort and energy, indesire and thinking. Experience morebasic than language tells us that if energyis expended effects are produced . We tendto believe that our bodies can stop upthis energy, prevent it from affecting otherthings until we will our bodies to overtaction. But this may be only because wehave our own linguistic basis for a theorythat formless items like 'matter' are thingsin themselves, malleable only by similarthings, by more matter, and hence insul-ated from the powers of life and thought .It is no more unnatural to think thatthought contacts everything and pervadesthe universe than to think, as we all do,that light kindled outdoors does this . And

'The Hopi verbs of preparing naturally donot correspond neatly to our 'prepare' ; so thatna'twani could also be rendered 'the practised-upon,' 'the tried-for,' and otherwise .

2 08

it is not unnatural to suppose that thought,like any other force, leaves everywheretraces of effect . Now when we think ofa certain actual rose-bush, we do not sup-pose that our thought goes to that actualbush, and engages with it, like a search-light turned upon it. What then do wesuppose our consciousness is dealing withwhen we are thinking of that rose-bush?Probably we think it is dealing with a'mental image' which is not the rose-bushbut a mental surrogate of it . But whyshould it be natural to think that ourthought deals with a surrogate and notwith the real rose-bush? Quite possiblybecause we are dimly aware that we carryabout with us a whole imaginary space,full of mental surrogates. To us, mentalsurrogates are old familiar fare. Alongwith the images of imaginary space, whichwe perhaps secretly know to be imaginaryonly, we tuck the thought-of actually ex-isting rose-bush, which may be quite an-other story, perhaps just because we havethat very convenient 'place for it. TheHopi thought-world has no imaginaryspace. The corollary to this is that it maynot locate thought dealing with real spaceanywhere but in real space, nor insulatereal space from the effects of thought. AHopi would naturally suppose that histhought (or he himself) traffics with theactual rose-bush- or more like, corn-plant -that he is thinking about. Thethought then should leave some trace ofitself with the plant in the field. If it isa good thought, one about health andgrowth, it is good for the plant ; if a badthought, the reverse .

The Hopi emphasize the intensity-factorof thought. Thought to be most effectiveshould be vivid in consciousness, definite,steady, sustained, charged with strongly-felt good intentions . They render the ideain English as 'concentrating,' 'hold it inyour heart,' 'putting your mind to it,'

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE

'earnestly hoping.' Thought power is theforce behind ceremonies, prayer-sticks,ritual smoking, etc. The prayer-pipe is re-garded as an aid to 'concentrating' (so

said my informant) . Its name, na'twanpi,means 'instrument of preparing.'

Covert participation is mental collabo-ration from people who do not take partin the actual affair, be it a job of work,hunt, race, or ceremony, but direct theirthought and good will toward the affair'ssuccess. Announcements often seek to en-list the support of such mental helpers aswell as of overt participants, and containexhortations to the people to aid withtheir active good will . 10 A similarity toour concepts of a sympathetic audience orthe cheering section at a football gameshould not obscure the fact that is is pri-marily the power of directed thought, andnot merely sympathy or encouragement,that is expected of covert participants . Infact these latter get in their deadliest workbefore, not during, the game! A corollaryto the power of thought is the power ofwrong thought for evil ; hence one pur-pose of covert participation is to obtainthe mass force of many good wishers tooffset the harmful thought of ill wishers .Such attitudes greatly favor cooperationand community spirit. Not that the Hopicommunity is not full of rivalries and col-liding interests . Against the tendency tosocial disintegration in such a small, iso-lated group, the theory of 'preparing' bythe power of thought, logically leadingto the great power of the combined, in-

10 See, e .g ., Ernest Beaglehole, Notes on HopiEconomic Life (Yale University Publications inAnthropology, No . 15, 1937), especially thereference to the announcement of a rabbit hunt,and on p. 30, description of the activities inconnection with the cleaning of Toreva Spring-announcing, various preparing activities, andfinally, preparing the continuity of the goodresults already obtained and the continued flowof the spring .

tensified and harmonized thought of thewhole community, must help vastlytoward the rather remarkable degree ofcooperation that in spite of much privatebickering the Hopi village displays in allthe important cultural activities .

Hopi 'preparing' activities again show aresult of their linguistic thought back-ground in an emphasis on persistence andconstant insistent repetition . A sense of thecumulative value of innumerable small mo-menta is dulled by an objectified, spatial-ized view of time like ours, enhanced bya way of thinking close to the subjectiveawareness of duration, of the ceaseless'latering' of events . To us, for whom timeis a motion on a space, unvarying repeti-tion seems to scatter its force along a rowof units of that space, and be wasted .To the Hopi, for whom time is not amotion but a 'getting later' of everythingthat has ever been done, unvarying repeti-tion is not wasted but accumulated . It isstoring up an invisible change that holdsover into later events." As we have seen,it is as if the return of the day were feltas the return of the same person, a littleolder but with all the impresses of yes-terday, not as 'another day,' i.e. like an

" This notion of storing up power, whichseems implied by much Hopi behavior, has ananalogue in physics, acceleration . It might besaid that the linguistic background of Hopithought equips it to recognize naturally thatforce manifests not as motion or velocity, butas cumulation or acceleration . Our linguisticbackground tends to hinder in us this samerecognition, for having legitimately conceivedforce to be that which produces change, wethen think of change by our linguistic meta-phorical analogue, motion, instead of by a puremotionless changingness concept, i.e., accumu-lation or acceleration. Hence it comes to ournaive feeling as a shock to find from physicalexperiments that it is not possible to defineforce by motion, that motion and speed, as also'being at rest,' are wholly relative, and thatforce can be measured only by acceleration.

209

ETC . : A REVIEW OF GENERAL SEMANTICS

entirely different person. This principlejoined with that of thought-power andwith traits of general Pueblo culture isexpressed in the theory of the Hopi cere-monial dance for furthering rain andcrops, as well as in its short, piston-liketread, repeated thousands of times, hourafter hour .

Some Impresses of Linguistic Habitin Western Civilization

It is harder to do justice in a few wordsto the linguistically-conditioned featuresof our own culture than in the case ofthe Hopi, because of both vast scope andand difficulty of objectivity-because ofour deeply ingrained familiarity with theattitudes to be analyzed . I wish merely tosketch certain characteristics adjusted toour linguistic binomialism of form plusformless item or 'substance,' to our meta-phoricalness, our imaginary space, and ourobjectified time. These, as we have seen,are linguistic.

From the form-plus-substance dichot-omy the philosophical views most tradi-tionally characteristic of the 'Westernworld' have derived huge support . Herebelong materialism, psycho-physical par-allelism, physics-at least in its traditionalNewtonian form-and dualistic views ofthe universe in general . Indeed here be-longs almost everything that is 'hard,practical common sense.' Monistic, hol-istic, and relativistic views of reality ap-peal to philosophers and some scientists,but they are badly handicapped for ap-pealing to the 'common sense' of theWestern average man . This is not becausenature herself refutes them (if she did,philosophers could have discovered thismuch) but because they must be talkedabout in what amounts to a new language .'Common sense,' as its name shows, and

210

'practicality' as its name does not show,are largely matters of talking so that oneis readily understood. It is sometimesstated that Newtonian space, time, andmatter are sensed by everyone intuitively,whereupon relativity is cited as showinghow mathematical analysis can prove in-tuition wrong. This, besides being unfairto intuition, is an attempt to answer off-hand question (1) put at the outset ofthis paper, to answer which this researchwas undertaken. Presentation of the find-ings now nears its end, and I think theanswer is clear . The offhand answer, lay-ing the blame upon intuition for our slow-ness in discovering mysteries of the cos-mos, such as relativity, is the wrong one .The right answer is: Newtonian space,time, and matter are no intuitions. Theyare recepts from culture and language .That is where Newton got them.

Our objectified view of time is how-ever favorable to historicity and to every-thing connected with the keeping of rec-ords, while the Hopi view is unfavorablethereto . The latter is too subtle, complex,and ever-developing, supplying no ready-made answer to the question of when'one' event ends and 'another' begins .When it is implicit that everything thatever happened still is, but is in a neces-sarily different form from what memoryor record reports, there is less incentiveto study the past. As for the present, theincentive would be not to record it butto treat it as 'preparing.' But our objecti-fied time puts before imagination some-thing like a ribbon or scroll marked offinto equal blank spaces, suggesting thateach be filled with an entry. Writing hasno doubt helped toward our linguistictreatment of time, even as the linguistictreatment has guided the uses of writing .Through this give-and-take between lan-guage and the whole culture we get, forinstance :

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE

1. Records, diaries, book-keeping, ac-counting, mathematics stimulated by ac-counting ;

2. Interest in exact sequence, dating,calendars, chronology, clocks, time wages,time graphs, time as used in physics ;

3 . Annals, histories, the historical atti-tude, interest in the past, archaeology, at-titudes of introjection towards past pe-riods, e.g., classicism, romanticism .

Just as we conceive our objectified timeas extending in the future like the wayit extends in the past, so we set down ourestimates of the future in the same shapeas our records of the past, producing pro-grams, schedules, budgets. The formalequality of the space-like units by whichwe measure and conceive time leads us toconsider the 'formless item' or 'substance'of time to be homogeneous and in ratio tothe number of units . Hence our prorataallocation of value to time, lending itselfto the building up of a commercial struc-ture based on time-prorata values : timewages (time work constantly supersedespiece work), rent, credit, interest, depre-ciation charges, and insurance premiums .No doubt this vast system once builtwould continue to run under any sort oflinguistic treatment of time ; but that itshould have been built at all, reaching themagnitude and particular form it has inthe Western world, is a fact decidedlyin consonance with the patterns of theSAE languages . Whether such a civiliza-tion as ours would be possible with widelydifferent linguistic handling of time is alarge question-in our civilization ourlinguistic patterns and the fitting of ourbehavior to the temporal order are whatthey are, and they are in accord . We areof course stimulated to use calendars,clocks, and watches, and to try to measuretime ever more precisely ; this aids science,and science in turn, following these well-worn cultural grooves, gives back to cul-

21 1

ture an ever-growing store of applications,habits, and values, with which cultureagain directs science. But what lies out-side this spiral? Science is beginning tofind that there is something in the cosmosthat is not in accord with the concepts wehave formed in mounting the spiral. It istrying to frame a new language by whichto adjust itself to a wider universe .

It is clear how the emphasis on 'savingtime' which goes with all the above andis very obvious objectification of time,leads to a high valuation of 'speed,' whichshows itself a great deal in our behavior .

Still another behavioral effect is that thecharacter of monotony and regularity pos-sessed by our image of time as an evenlyscaled limitless tape measure persuades usto behave as if that monotony were moretrue of events than it really is . That is, ithelps to routinize us. We tend to selectand favor whatever bears out this view,to 'play up to' the routine aspects of exist-ence. One phase of this is behavior evinc-ing a false sense of security or an assump-tion that all will always go smoothly, anda lack in foreseeing and protecting our-selves against hazards. Our technique ofharnessing energy does well in routineperformance, and it is along routine linesthat we chiefly strive to improve it-weare, for example, relatively uninterestedin stopping the energy from causing acci-dents, fires, and explosions, which it isdoing constantly on a wide scale. Such in-difference to the unexpectedness of lifewould be disastrous to a society as small,isolated, and precariously poised as theHopi society is, or rather once was .

Thus our linguistically-determinedthought world not only collaborates withour cultural idols and ideals, but engageseven our unconscious personal reactions inits patterns and gives them certain typicalcharacters . One such character, as we haveseen, is carelessness, as in reckless driving

ETC. : A REVIEW OF GENERAL SEMANTICS

or throwing cigarette stubs into wastepaper. Another of different sort is gestur-ing when we talk. Very many of the ges-tures made by English-speaking people atleast, and probably by all SAE speakers,serve to illustrate by a movement in space,not a real spatial reference but one of thenon-spatial references that our languagehandles by metaphors of imaginary space .That is, we are more apt to make a grasp-ing gesture when we speak of graspingan elusive idea than when we speak ofgrasping a doorknob. The gesture seeks tomake a metaphorical and hence somewhatunclear reference more clear . But if a lan-guage refers to non-spatials without im-plying a spatial analogy, the reference isnot made any clearer by gesture . The Hopigesture very little, perhaps not at all in thesense we understand as gesture .

It would seem as if kinesthesia, or thesensing of muscular movement, thougharising prior to language, should be mademore highly conscious by linguistic use ofimaginary space and methaphorical im-ages of motion. Kinesthesia is marked intwo facets of European culture : art andsport. European sculpture, an art in whichEurope excels, is strongly kinesthetic, con-veying great sense of the body's motions ;European painting likewise . The dance inour culture expresses delight in motionrather than symbolism or ceremonial, andour music is greatly influenced by ourdance forms. Our sports are strongly im-bued with this element of the 'poetry ofmotion.' Hopi races and games seem toemphasize rather the virtues of enduranceand sustained intensity. Hopi dancing ishighly symbolic and is performed withgreat intensity and earnestness, but hasnot much movement or swing.

Synesthesia, or suggestion by certainsense receptions of characters belongingto another sense, as of light and color bysounds and vice versa, should be made

212

more conscious by a linguistic metaphor-ical system that refers to non-spatial ex-periences by terms for spatial ones, thoughundoubtedly it arises from a deepersource. Probably in the first instance meta-phor arises from synesthesia and not thereverse, yet metaphor need not becomefirmly rooted in linguistic pattern, as Hopishows. Non-spatial experience has onewell-organized sense, hearing-for smelland taste are but little organized . Non-spatial consciousness is a realm chiefly ofthought, feeling, and sound . Spatial con-sciousness is a realm of light, color, sight,and touch, and presents shapes and di-mensions . Our metaphorical system, bynaming non-spatial experiences after spa-tial ones, imputes to sounds, smells, tastes,emotions, and thoughts qualities like thecolors, luminosities, shapes, angles, tex-tures, and motions of spatial experience.And to some extent the reverse transfer-ence occurs ; for after much talking abouttones as high, low, sharp, dull, heavy,brilliant, slow, the talker finds it easy tothink of some factors in spatial experi-ence as like factors of tone . Thus wespeak of 'tones' of color, a gray 'mono-tone,' a 'loud' necktie, a 'taste' in dress ;all spatial metaphor in reverse. NowEuropean art is distinctive in the way itseeks deliberately to play with synesthesia .Music tries to suggest scenes, color, move-ment, geometric design ; painting andsculpture are often consciously guided bythe analogies of music's rhythm ; colorsare conjoined with feeling for the analogyto concords and discords . The Europeantheatre and opera seek a synthesis ofmany arts. It may be that in this way ourmetaphorical language that is in somesense a confusion of thought is producing,through art, a result of far-reaching value-a deeper esthetic sense leading towarda more direct apprehension of underlyingunity behind the phenomena so variouslyreported by our sense channels .

autocratic way. This is because a languageis a system, not just an assemblage ofnorms. Large systemic outlines can changeto something really new only very slowly,while many other cultural innovations aremade with comparative quickness. Lan-guage thus represents the mass mind ; itis affected by inventions and innovations,but affected little and slowly, whereas toinventors and innovators it legislates withthe decree immediate.

The growth of the SAE language-cul-ture complex dates from ancient times .Much of its metaphorical reference to thenon-spatial by the spatial was already fixedin the ancient tongues, and more espe-cially in Latin . It is indeed a marked traitof Latin. If we compare, say Hebrew, wefind that while Hebrew has some allusionto not-space as space, Latin has more .Latin terms for non-spatials, like educo,religio, principia, comprehendo, are usu-ally metaphorized physical references :lead out, tying back, etc . This is not trueof all languages-it is quite untrue ofHopi. The fact that in Latin the directionof development happened to be from spa-tial to non-spatial (partly because of sec-ondary stimulation to abstract thinkingwhen the intellectually crude Romans en-countered Greek culture) and that latertongues were strongly stimulated to mimicLatin, seems a likely reason for a beliefwhich still lingers on among linguists thatthis is the natural direction of semantic

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE

Historical Implications

change in all languages, and for the per-sistent notion in Western learned circles

How does such a network of language, (in strong contrast to Eastern ones) thatculture, and behavior come about his- objective experience is prior to subjective .torically? Which eras first, the language Philosophies make out a weighty case forpatterns or the cultural norms? In main the reverse, and certainly the direction ofthey have grown up together, constantly development is sometimes the reverse.influencing each other . But in this part- Thus the Hopi word for 'heart' can benership the nature of the language is the shown to be a late formation within Hopifactor that limits free plasticity and rigid- from a root meaning think or remember.ifies channels of development in the more Or consider what has happened to the

word 'radio' in such a sentence as 'hebought a new radio,' as compared to itsprior meaning 'science of wireless tele-phony .'

In the middle ages the patterns alreadyformed in Latin began to interweave withthe increased mechanical invention, indus-try, trade, and scholastic and scientificthought . The need for measurement in in-dustry and trade, the stores and bulks of'stuffs' in various containers, the type-bodies in which various goods were han-dled, standardizing of measure and weightunits, invention of clocks and measure-ment of 'time,' keeping of records, ac-counts, chronicles, histories, growth ofmathematics and the partnership ofmathematics and science, all cooperated tobring our thought and language worldinto its present form .

In Hopi history, could we read it, weshould find a different type of languageand a different set of cultural and en-vironmental influences working together .A peaceful agricultural society isolated bygeographic features and nomad enemiesin a land of scanty rainfall, arid agricul-ture that could be made successful onlyby the utmost perseverance (hence thevalue of persistence and repetition), neces-sity for collaboration (hence emphasis onthe psychology of teamwork and on men-tal factors in general), corn and rain asprimary criteria of value, need of exten-sive preparations and precautions to as-

2 1 3

ETC . : A REVIEW OF GENERAL SEMANTICS

sure crops in the poor soil and precariousclimate, keen realization of dependenceupon nature favoring prayer and a religi-ous attitude toward the forces of nature,especially prayer and religion directedtoward the ever-needed blessing, rain-these things interacted with Hopi lin-guistic patterns to mold them, to bemolded again by them, and so little bylittle to shape the Hopi world-outlook.

To sum up the matter, our first ques-tion asked in the beginning (p . 200) isanswered thus : Concepts of 'time' and'matter' are not given in substantially thesame form by experience to all men butdepend upon the nature of the languageor languages through the use of whichthey have been developed. They do notdepend so much upon any one system(e.g ., tense, or nouns) within the gram-mar as upon the ways of analyzing andreporting experience which have becomefixed in the language as integrated'fashions of speaking' and which cutacross the typical grammatical classifica-tions, so that such a 'fashion' may includelexical, morphological, syntactic, andotherwise systematically diverse means co-ordinated in a certain frame of consist-ency. Our own 'time' differs markedlyfrom Hopi 'duration.' It is conceived aslike a space of strictly limited dimensions,or sometimes as like a motion upon sucha space, and employed as an intellectualtool accordingly . Hopi 'duration' seems tobe inconceivable in terms of space ormotion, being the mode in which life dif-fers from form, and consciousness in totofrom the spatial elements of conscious-ness. Certain ideas born of our own time-concept, such as that of absolute simul-taneity, would be either very difficult toexpress or impossible and devoid of mean-ing under the Hopi conception, and wouldbe replaced by operational concepts. Our'matter' is the physical sub-type of 'sub-

2 14

stance' or 'stuff,' which is conceived as theformless extensional item that must bejoined with form before there can be realexistence. In Hopi there seems to benothing corresponding to it ; there are noformless extensional items ; existence mayor may not have form, but what it alsohas, with or without form, is intensityand duration, these being non-extensionaland at bottom the same .

But what about our concept of 'space,'which was also included in our first ques-tions? There is no such striking differencebetween Hopi and SAE about space asabout time, and probably the apprehensionof space is given in substantially the sameform by experience irrespective of lan-guage. The experiments of the Gestaltpsychologists with visual perception ap-pear to establish this as a fact. But theconcept o f space will vary somewhat withlanguage, because as an intellectual toohzit is so closely linked with the concomitantemployment of other intellectual tools, ofthe order of 'time' and 'matter,' whichare linguistically conditioned. We seethings with our eyes in the same spaceforms as the Hopi, but our idea of spacehas also the property of acting as a surro-gate of non-spatial relationships like time,intensity, tendency, and as a void to befilled with imagined formless items, oneof which may even be called 'space .'Space as sensed by the Hopi would notbe connected mentally with such surro-gates, but would be comparatively 'pure,'unmixed with extraneous notions .

As for our second question (p . 201)There are connections but not correlationsor diagnostic correspondences betweencultural norms and linguistic patterns .Although it would be impossible to inferthe existence of Crier Chiefs from thelack of tenses in Hopi, or vice versa, there

u Here belong 'Newtonian' and 'Euclidean'space, etc.

RELATION OF HABITUAL THOUGHT AND BEHAVIOR TO LANGUAGE .

is a relation between a language and therest of the culture of the society whichuses it. There are cases where the `fashionsof speaking' are closely integrated with thewhole general culture, whether or not thisbe universally true, and there are connec-tions within this integration, between thekind of linguistic analyses employed andvarious behavioral reactions and also theshapes taken by various cultural develop-ments. Thus the importance of CrierChiefs does have a connection, not withtenselessness itself, but with a system of

thought in which categories different fromour tenses are natural. These connectionsare to be found not so much by focusingattention on the typical rubrics of lin-guistic, ethnographic, or sociological de-scription as by examining the culture andthe language (always and only when thetwo have been together historically for aconsiderable time) as a whole in whichconcatenations that run across these de-partmental lines may be expected to exist,and if they do exist, eventually to be dis-coverable by study .

A NOTE ON WORD-MAGIC[Joseph, speaking of a lion .] `But if he had come, with lashing tail,

and roared after his prey, like the voice of the chanting seraphim, yetthy child would have been little affrighted or not at all before his rage .. . . For knoweth not my father that the beasts fear and avoid man, forthat God gave him the spirit of understanding and taught him theorders into which single things fall ; doth he not know how Shemmaelshrieked when the man of earth knew how to name the creation asthough he were its master and framer . . . ? And the beasts too theyare ashamed and put the tail between their legs because we know themand have power over their names and can thus render powerless theroaring might of the single one, by naming him. If now he had come,with long slinking tread, with his hateful nose, mewing and spitting,terror would not have robbed me of my senses, nor made me palebefore his riddle. "Is thy name Blood-Thirst?" I would have asked ofhim, making merry at his expense . "Or Springing Murder?" But thereI would have sat upright and cried out : "Lion! Lo, Lion art thou, bynature and species, and thy riddle lieth bare before me, so that I speakit out and with a laugh it is plain ." And he would have blinked beforethe name and gone meekly away before the word, powerless to answerunto me. For he is quite unlearned and knows nought of writingtools .'

THOMAS MANN, Joseph and His Brethren

21 5