structure foreign news 1965 - galtung

8/3/2019 Structure Foreign News 1965 - Galtung

1/29

The Structure of Foreign NewsAuthor(s): Johan Galtung and Mari Holmboe RugeSource: Journal of Peace Research, Vol. 2, No. 1 (1965), pp. 64-91Published by: Sage Publications, Ltd.Stable URL: http://www.jstor.org/stable/423011

Accessed: 01/10/2009 05:05

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

http://www.jstor.org/action/showPublisher?publisherCode=sageltd.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with thescholarly community to preserve their work and the materials they rely upon, and to build a common research platform that

promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

Sage Publications, Ltd. is collaborating with JSTOR to digitize, preserve and extend access toJournal of Peace

Research.

http://www.jstor.org
http://www.jstor.org/stable/423011?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=sageltdhttp://www.jstor.org/action/showPublisher?publisherCode=sageltdhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/423011?origin=JSTOR-pdf


2/29

THE STRUCTURE OF FOREIGN NEWSThe Presentation of the Congo, Cuba and Cyprus Crises

in Four Norwegian Newspapers*By

JOHAN GALTUNG and MARI HOLMBOE RUGEPeace Research Institute - Oslo

1. IntroductionIn this article the general problem offactors influencing the flow of news fromabroad will be discussed, following thekind of reasoning given by Ostgaard inhis article,l but in a somewhat differentway. A systematic presentation of factorsthat seem to be particularlyimportant willbe followed by a simple theory and thededuction of some hypotheses from them.No claim is made for completeness inthe list of factors or 'deductions'. Some ofthese hypotheses will then be tested ondata relating to the presentation in fourNorwegian newspapers of three parti-cular and recent crises abroad. Gaps inour present knowledge will be indicatedand some possible policy implicationsdrawn.The point of departure is our world asa geographic structure divided roughlyinto 160 territories, most of which arecalled nations and are 'autonomous'. Theinternational community of nations isstructured by a number of variables andhighly stratified into 'topdog' and 'under-dog' nations so that the world is geographyon which are superimposed two rela-tively similar levels of human organiza-tion: the inter-individual and the inter-national. The twvolevels are not inde-pendent of each other and the more theyare linked (the more population andleadership in any nation are interdepen-dent), and the more nations are inter-dependent because of increasing efficiencyof communication and militalry action,2

the more valid is the old sociologicalslogan about 'everything's relevance foreverything else'.Thus, the world consists of individualand national actors, and since it is axio-matic that action is based on the actor'simage of reality, international action willbe based on the image of internationalreality. This image is not shaped by thenews media (press, radio, TV, newsreels)alone; personal impressions and contacts,professional relations abroad, diplomaticdispatches, etc., count too - whetherless, equally much or more, we do notknow. But the regularity, ubiquity andperseverence of news media will in anycase make them first-rate competitors forthe number-one position as internationalimage-former. Since the adequacy of anaction is often, but by no means always,positively related to the adequacy of theimage on which it is based,3 research intothe adequacy of the image the newsmedia give of the world is of primaryimportance.At the interpersonal level the relation-ship between the events, the perceptionwithall the selective and distorting factors thatare operative under different circum-stances, and the final image is relativelywell explored. At the level of collectiveperception, where perception is made onbehalf of others to be relayed to theseothers later, the situation is much morecomplicated. From world events to per-sonal image we have the chain of com-munication presented in Figure 1.


3/29

65Figure 1. The chainof newscommunication

world events -> media -+ media - personal -- personalperception t image perception tf imageselectiondistortionWe are concerned with the first half

of this chain, from world events to newsimage, or, to be more specific, to theprinted page in the newspaper since ourdata refer to that. In other words: how do'events' become 'news'? This does not meanthat the second half is unimportant - onthe contrary, it is the personal image, notthe newspaper that counts, but this willbe discussed in a later article. In analyzingthe first half we shall treat the news mediaas non-personal indivisible entities andnot distinguish between the journalist inthe field in the news-sending country, thelocal press agency bureau, the districtbureau, the central bureau of the pressagency, the district bureau on the receiv-ing end, the local bureau in the news-receiving country, the news editor in thereceiving newspaper, the layout man, andwhat not - to indicate a chain with someseven or eight steps in it.4 The chain mayof course be much shorter if the news-paper has a correspondent; it may thenbe reduced to event-correspondent-editor,which involves two steps only. Ostgaardhas indicated many of the problems alongthis chain5 and detailed analysis here iscertainly important for future research,but our analysis will treat news media inabstractoand limit itself to some reasoningfrom first principles.2. The theoryTo do this a metaphor with sufficientheuristic power to offer insights (but cer-tainly not proofs) is useful. One suchmetaphor is as follows. Imagine that theworld can be likened to an enormous setof broadcasting stations, each one emittingits signal or its program at its properwavelength. (Another metaphor might beof a set of atoms of different kinds emittingwaves corresponding to their condition.)

selectiondistortionThe emission is continuous, correspondingto the truism that something is always hap-

pening to any person in the world. Evenif he sleeps quietly, sleep is 'happening'6- what we choose to consider an 'event'is culturally determined. The set of worldevents, then, is like the cacophony ofsound one gets by scanning the dial ofone's radio receiver, and particularly con-fusing if this is done quickly on the medium-wave or short-wave dials. Obviously thiscacophony does not make sense, it maybecome meaningful only if one stationis tuned in and listened to for sometime before one switches on to the nextone.

Since we cannot register everything, wehave to select, and the question is what willstrike our attention. This is a problemin the psychology of perception and thefollowing is a short list of some obviousimplications of this metaphor:F1: If the frequency of the signal is out-side the dial it will not be recorded.F2: The strongerthe signal, the greatertheamplitude, the more probable that itwill be recordedas worth listening to.F3: The more clear and unambiguous thesignal (the less noise there is), themoreprobable that it will be recordedas worth listening to.F4: The more meaningful the signal, themoreprobable that it will be recordedas worth listening to.F5: The moreconsonantthe signal is withthe mental image of what one expectstofind, the moreprobable that it willbe recordedas worth listening to.FB: The more unexpected the signal, themoreprobable that it will be recordedas worth listening to.F7: If one signal has been tuned in to themorelikely it will continueto be tuned

in to as worth listening to.F8: The morea signal has beentunedin to,the moreprobable that a very diffferentkind of signal will be recorded asworth listening to next time.


4/29

66Some comments on these factors are in

order. They are nothing but common-sense perception psychology translatedinto radio-scanning and event-scanningactivities. The proper thing to do in orderto test their validity would be to observejournalists at work or radio listeners oper-ating with the dial - and we have nosuch data. For want of this the factorsshould be anchored in general reasoningand social science findings (but referencesto the latter will be given in the notes onlysince they are not essential to our reason-ing).

The first factor is trivial when appliedto radio sets, less so when applied toevents in general. Since this is a metaphorand not a model we shall be liberal in ourinterpretation of 'frequency' and proceedas follows. By the 'frequency' of an eventwe refer to the time-span needed for theevent to unfold itself and acquire meaning.For a soldier to die during a battle thistime-span is very short; for a developmentprocess in a country to take place thetime-span may be very long. Just as theradio dial has its limitation with regard toelectro-magnetic waves, so will the news-paper have its limitations, and the thesisis that the more similar the frequency of theevent is to thefrequency of the news medium,themoreprobablethat it will be recorded s newsby that news medium. A murder takes littletime and the event takes place between thepublication of two successive issues of adaily, which means that a meaningfulstory can be told from one day to the next.But to single out one murder during abattle where there is one person killedevery minute would make little sense -one will typically only record the battleas such (if newspapers were publishedevery minute the perspective could pos-sibly be changed to the individual soldier).Correspondingly, the event that takesplace over a longer time-span will go un-recorded unless it reaches some kind ofdramatic climax (the building of a damgoes unnoticed but not its inauguration).

Needless to say, this under-reporting oftrends is to some extent corrected by publi-cations with a lower frequency. A news-paper may have a habit of producingweekly 'reviews', there are weeklies andmonthlies and quarterlies and* yearbooks- and there are ad hoc publications. Ifwe concentrate on dailies, however, thethesis is probably valid and probably ofsome heuristic value when other aspects ofnews communication are to be unraveled.

The second thesis is simply that there issomething corresponding to the idea of'amplitude' for radio waves. What thissays is only that the bigger the dam, themore will its inauguration be reportedceterisparibus; the more violent the murderthe bigger the headlines it will make. Itsays nothing about what has greater am-plitude, the dam or the murder. It can alsobe put in a more dichotomous form: thereis a threshold the event will have to passbefore it will be recorded at all.7 This isa truism, but an important one.The third hypothesis is also trivial atthe radio level but not at the news level.What is 'signal' and what is 'noise' is notinherent; it is a question of convention,8as seen clearly when two radio stations aresending on the same frequency. Clarityin this connection must refer to some kindof one-dimensionality, that there is onlyone or a limited number of meanings inwhat is received. Thus interpreted thehypothesis says simply the following: theless ambiguity the more the event will benoticed. This is not quite the same as pre-ferring the simple to the complex, butone precization of it rather; an event witha clear interpretation, free from ambi-guities in its meaning, is preferred to thehighly ambiguous event from which manyand inconsistent implications can andwill be made.9

The fourth hypothesis also deals withmeaning but not with its ambiguity.'Meaningful' has some major interpreta-tions. One of them is 'interpretable withinthe cultural framework of the listener or


5/29

reader' and all the thesis says is that actu-ally some measure of ethnocentrismwill beoperative: there has to be cultural proxi-mity. That is, the event-scanner will payparticular attention to the familiar, to theculturally similar, and the culturallydistant will be passed by more easily andnot be noticed. It is somewhat like theNorth European radio listener in, say,Morocco: he will probably pass by theArab music and speech he can get on hisdial as quaint and meaningless and findrelief in European music and French talk.The other dimension of 'meaningful' isin terms of relevance:an event may happenin a culturally distant place but still beloaded with meaning in terms of what itmay imply for the reader or listener. Thusthe culturally remote country may bebrought in via a pattern of conflict withone's own group.10The fifth hypothesis links what is selec-ted to the mental pre-image, where theword 'expects' can and should be givenboth its cognitive interpretation as 'pre-dicts' and its normative interpretation as'wants'. A person predicts that somethingwill happen and this creates a mentalmatrix for easy reception and registrationof the event if it does finally take place.Or he wants it to happen and the matrixis even more prepared, so much so thathe may distort perceptions he receivesand provide himself with images conso-nant with what he has wanted. In the sensementioned here 'news' are actually 'olds',because they correspond to what oneexpects to happen - and if they are toofar away from the expectation they willnot be registered, according to this hypo-thesis of consonance.,lThe sixth hypothesis brings in a correc-tive to the fourth and fifth. The idea issimply that it is not enough for an eventto be culturally meaningful and consonantwith what it expected - this defines onlya vast set of possible news candidates.Within this set, according to the hypothe-sis, the more unexpected have the highest

67chances of being included as news. It isthe unexpected within the meaningful andthe consonant that is brought to one's at-tention, and by 'unexpected' we simplymean essentially two things: unexpectedorrare. Thus, what is regular and institu-tionalized, continuing and repetitive atregular and short intervals does not attractnearly so much attention, ceterisparibus, asthe unexpected and ad hoc - a circum-stance that is probably well known to theplanners of summit meetings.l2 Events haveto be unexpected or rare, or preferablyboth, to become good news.

The seventh hypothesis is the idea thatonce something has hit the headlines andbeen defined as 'news', then it will continueto be defined as news for some time evenif the amplitude is drastically reduced.l3The channel has been opened and stayspartly open to justify its being opened inthe first place, partly because of inertia inthe system and partly because what wasunexpected has now also become familiar.Thus F7 is, in a sense, deducible from F3and F6.

The eighth and final hypothesis refers tothe composition of such units as eveningentertainment for the family around theradio set, the front page of a newspaper,the newscast on radio, the newsreel on TVor in the cinema, and so on. The idea isthis: imagine the news editor of a broad-casting station has received only news fromabroad and only of a certain type. Someminutes before he is on the air he getssome insignificant domestic news and someforeign news of a different kind. The hypo-thesis is that the threshold value forthese news items will be much lower thanwould otherwise have been the case, be-cause of a desire to present a 'balanced'whole. Correspondingly, if there are al-ready many foreign news items the thresh-old value for a new item will be increased.

As mentioned, these eight factors arebased on fairly simple reasoning aboutwhat facilitates and what impedes percep-tion. They are held to be culture-free in


6/29

68the sense that we do not expect them tovary significantly with variations in hu-man culture - they should not dependmuch on cultural parameters. More par-ticularly,we would not expect them to varymuch along the east-west, north-south orcenter-peripheryaxes which we often makeuse of to structure the world. In particular,these factors should be relatively inde-pendent of some other major determinantsof the press. A newspaper may vary in thedegree to which it caters to mass circula-tion and a freemarket economy. If it wantsa mass circulation, all steps in the newschain will probably anticipate the reactionof the next step in the chain and accen-tuate the selection and distortion effectsin order to make the material more com-patiblewith theirimage of what the readerswant. Moreover, a newspaper may varyin the degree to which it tries to presentmany aspects of the situation, or, rather,like the partners in a court case, try topresent only the material that is easilycompatible with its own political pointof view. In the latter case selection anddistortion will probably be accentuatedand certainly not decrease.But there is little doubt that there arealso culture-bound factors influencing thetransition from events to news, and weshall mention four such factors that wedeem to be important at least in the north-western corner of the world. They are:

Fg: The morethe eventconcernslite na-tions, the moreprobable hat it willbecome newsitem.F10: Themoreheevent oncernslitepeople,themoreprobable hat it will becomea newsitem.F11: Themore heevent anbe seen nper-sonal terms,as due to the actionofspecific ndividuals,hemoreprobablethatit will become newsitem.F12: The morenegativeheevent n its con-sequences,he moreprobablethat itwill become newsitem.Again, some comments are in order.That news is elite-centered,n terms ofnations or in terms of people, is hardly

strange. The actions of the elite are, atleast usually and in short-term perspec-tive, more consequential than the activi-ties of others: this applies to elite nationsas well as to elite people. Moreover, asamply demonstrated by the popular maga-zines found in most countries, the elite canbe used in a sense to tell about everybody.A story about how the king celebrates hisbirthday will contain many elements thatcould just as well have been told aboutanybody, but who in particular amongordinarymen and women should be pickedfor the telling of the story? Elite peopleare available to serve as objects of generalidentification, not only because of theirintrinsic importance. Thus in an elite-centered news communication systemordi-nary people are not even given the chanceof representing themselves. Mutatis mu-tandis,the same should apply to nations.More problematic is the idea of personi-fication. The thesis is that news has a ten-dency to present events as sentenceswhere there is a subject, a named personor collectivity consisting of a few persons,and the event is then seen as a consequenceof the actions of this person or these per-sons. The alternative would be to presentevents as the outcome of 'social forces',as structural more than idiosyncratic out-comes of the society which produced them.In a structural presentation the names ofthe actors would disappear much asthey do in sociological analysis and muchfor the same reason - the thesis is that thepresentation actually found is more similarto what one finds in traditional personifiedhistorical analysis.To the extent that this isthe case the problem is why, and we havefive different explanations to offer:

1. Personification is an outcome ofcultural dealismaccording to whichman is the master of his own des-tiny and events can be seen as theoutcome of an act of free will. Ina culture with a more materialisticoutlook this should not be the case.Structural factors should be em-phasized, there will be more events


7/29

happening to people or with peopleas instruments than events causedby people.2. Personificationis a consequence ofthe need for meaning and conse-quently for identification: personscan serve more easily as objects ofpositive and negative identifica-tion through a combination ofprojection and empathy.3. Personification is an outcome ofthe frequency-factor: persons canact during a time-span that fitsthe frequency of the news media,'structures' are more difficult topin down in time and space.4. Personification can be seen as adirect consequence of the elite-concentrationut as distinct from it.5. Personification is more in agree-ment with modern techniques ofnews gathering and news presenta-tion. Thus, it is easier to take aphoto of a person than of a 'struc-ture' (the latter is better formovies- perhaps), and whereas oneinterview yields a necessary andsufficient basis for one person-cen-tered news story, a structure-cen-tered news storywill require manyinterviews,observationtechniques,data gathering, etc. Obviouslythere is an egg-chicken argumentimplied here since it may also beargued that personification camefirst and that techniques,the wholestructure of news communication,were developed accordingly.

We only offer those explanations with-out choosing between them; first of allbecause there is no reason to choose aslong as they do not contradict each other,and secondly because we have neitherdata nor theory that can provide us witha rational basisfor a choice. It is our hunchthat future research will emphasize thatthese factors reinforce each other in pro-ducing personification.When we claim that negativenews willbe preferred to positive news we aresaying nothing more sophisticated thanwhat most people seem to refer to whenthey say that 'there is so little to be happyabout in the news', etc. But we can offera number of reasons why this state of af-

69fairs appears likely, just as we did for thefactor of personification. We shall do sousing the other factors relatively systemat-ically:

1. Negative news enters the newschannel more easily because itsatisfies the frequency riterion bet-ter. There is a basic asymmetryinlife between the positive, whichis difficult and takes time, and thenegative, which is much easier andtakes less time - compare theamount of time needed to bringup and socialize an adult personand the amount of time needed tokill him in an accident: the amountof time needed to build a houseand to destroy it in a fire, to makean airplane and to crash it, andso on. The positive cannot be tooeasy, for then it would have lowscarcity value. Thus, a negativeevent can more easily unfold it-self completely between two issuesof a newspaper and two newscasttransmissions - for a positiveevent this is more difficult andspecific. Inaugurating or culmi-nating events are needed. A P.R.-minded operator will, of course,see to that - but he is not alwayspresent.2. Negative news will more easily beconsensual and unambiguous in thesense that there will be agreementabout the interpretation of theevent as negative. A 'positive'event may be positive to somepeople and not to others andhence not satisfy the criterion ofunambiguity. Its meaning will beblurred by other overtones andundertones.3. Negative news is said to be moreconsonant ith at least some domi-nant pre-images of our time. Theidea must be that negative newsfulfills some latent or manifestneeds and that many people havesuch needs. Of the many theoriesin this field we prefer the cognitivedissonance version because it isfalsifiable. The theory, however,presupposesa relatively high levelof general anxiety to provide asufficient matrix in which nega-tive news can be embedded withmuch consonance. This should be


8/29

70the case during crises,l4 so a testof this theory would be that dur-ing crises news that is not relatedto the crisis tends to be more nega-tive and not more positive (as atheory of compensation ratherthan of dissonance/reduction wouldpredict).4. Negative news is more unexpectedthan positive news, both in thesense that the events referred to aremore rare, and in the sense thatthey are less predictable. This pre-supposes a culture where changesto the positive, in other words'progress', are somehow regardedas the normal and trivial thingthat can pass under-reported be-cause it represents nothing new.The negative curls and eddiesrather than the steady positiveflow will be reported. The test ofthis theory would be a culturewith regressas the normal, and inthat case one would predict over-reporting of positive news. Thisis exemplified by news about theillness of an important person: theslightest improvement is over-re-ported relative to a steady decline.

Again we do not have sufficient theory tomake a choice between these possible ex-planations - nor do we have to do sosince they do not exclude each other.As to these last four factors it was men-tioned that they seem to be of particularimportance in the northwestern corner ofthe world. This does not mean that theyare not operating in other areas, but onecould also imagine other patterns of rela-tionship between the set of events and theset of news. Table 1 shows some examples:

Pattern I is the pattern we have de-scribed above. Pattern II would, where thelast two aspects are concerned, be more inagreement with socialist thinking, andwhere the first two are concerned, withbig-power thinking. It might fit the newsstructure of the Soviet Union, but withthe important proviso that one wouldprobably use pattern III to describeWestern powers. Similarly, a newly in-dependent developing nation might usepattern IV for itself, but also reserve pat-tern III for former colonial powers. Butall this is very speculative.'5

Let us then list systematically the twelvefactors we have concentrated on in thisanalysis; with subfactors:Events become news to the extent thatthey satisfy the conditions of

F,: frequencyF2: thresholdF21: absolute intensityF.22: intensity increaseF3: unambiguityF4: meaningfulnessF4.,: culturalproximityF4.2: relevanceFg: consonanceF5,: predictabilityF5 2: demandF6: unexpectednessF6.,: unpredictabilityF,.2: scarcityF,: continuityF8: compositionFg: referenceto elite nations

Table 1. Somepatternsof news structureF9 F1o F,, F,2Pattern nation people personification negativization

I .............. elite elite person negativecentered centered centered centeredII .............. elite elite structure positivecentered centered centered centered

III .............. elite elite both negativecentered centered centeredIV .............. non-elite elite person positivecentered centered centered centered


9/29

F,0: reference o elite peopleF,,: referenceto personsF12: referenceto something negative

As mentioned, these twelve factors arenot independent of each other: there areinteresting inter-relations between them.However, we shall not attempt to 'axio-matize' on this meager basis.Let us now imagine that all these fac-tors are operating. This means, we hypo-thesize, three things:1. The more events satisfy the criteria men-

tioned, the more likely that they will beregisteredas news (selection).2. Once a news item has been selected whatmakes it newsworthyaccordingto thefactorswill be accentuated distortion).3. Both the processof selection and theprocessof distortion will take place at all steps inthe chain rom eventto reader(replication).

Thus the longer the chain, the moreselection and distortion will take place ac-cording to this - but the more materialwill there also be to select from and todistort if one thinks of the press agenciesrelative to special correspondents. Inother words, we hypothesize that everylink in the chain reacts to what it receivesfairly much according to the same prin-ciples. The journalistscans the phenomena(in practice to a large extent by scanningother newspapers)and selects and distorts,and so does the reader when he gets thefinished product, the news pages, and sodo all the middle-men. And so do, weassume, people in general when they reportsomething, and, for instance, diplomatswhen they gather material for a dispatchto their ministry - partly because theyare conditioned by their psychology andtheir culture, partly because this is re-inforced by the newspapers.In general this means that the cumula-tive effects of the factors should be con-siderable and produce an image of theworld different from 'what really hap-pened' - for instance in the ways indi-

71cated by Ostgaard.16 However, since wehave no base-line in direct reports on'what really happened' on which this canbe tested we shall proceed in a differentdirection. Our problem is how the factorsrelate to each other in producing a finaloutcome.

Imagine that all factors, for the sake ofsimplicity, are dichotomized so that anevent either possesses them or does notpossess them. A given event can receive ascore from 0-12 according to this system,and we claim that this is as good a scoreof that elusive concept of 'newsworthi-ness' as any, in a culture where F8-F12are valid. This has two theoretical impli-cations that will be spelt out. The first oneis almost too simple to mention:Additivity hypothesis: The higher the totalscore of an event, the higher the probabilitythat it will become news, and even makeheadlines.

This may be seen as a hypothesis abouthow journalists work, about how the nighteditor reacts to incoming news-script orabout how the reader reacts when hescans his newspaper for something worthreading. It may be more valid in the firsttwo than in the last case - we do notknow. But it is interesting to put downsome pairs that should be considered parti-cularly newsworthy:( 9, 10): news about elite people in elite

nationsl7( 9, 12): news of a negative nature relatingto elite nations - in other words,big power conflict

(10, 12): news of a negativenature relating toelitepeople in otherwords, strugglefor power etc., at the top of society(11, 12): news of a negativenature relating topersons - in otherwords, scandals.

It is hardly necessary to make a contentanalysis to substantiate the claim thatthese four categories account for a sizeablefraction of the news presented by news-papers in most parts of the world.


10/29

72But there is another hypothesis that is a high F; has already contributed towardsless trivial. An event obviously does not the total score. According to the additivityhave to score 12 to hit the headlines. hypothesis there will also be news whereImagine the floor level for acceptance is both are high, and much prominence willat score 6, which can be obtained in (162) be given to them. But events where both= 924 differentways. (This high number, are low will not be admitted as news.by the way, explains why factors may be Thus, for the simple case of two factorsoperating and still not be noticed by the only, Fi and Fj, we have the three kindspublic: the variety is too great). The im- of events indicated in Table 2:

Table 2. A trichotomyf eventsaccordingo newsworthinessFi Fj Scoreof news-worthiness

Type 1. Prominent ews ........ .............. high high 2Type 2. Ordinary........................ high low 1low highType 3. Events,not news .................. low low Oplication of this is only that if the eventis low on one dimension or factor it maycompensatcor that by being high on another,and still make the news. For instance, theless an event refers to personsas actors themore negative will it have to be (earth-quakes, accidents that are presented interms of technical errors, not in terms of'the human factor'). The more culturallyclose and hence meaningful the event, theless does it have to refer to elite people -and vice versa: the more culturally distantthe event, the more should it refer to elitepeople, ceterisparibus (which correspondsto the impressionthat rank-and-filepeopleare highly under-reported when they livein far away countries). And so on, and soforth: this will be spelt out in section 5below:Since we have 12 factors this principlegives rise to (122) = 66 hypotheses, all ofthe following form:Complementarityhypothesis:

Fi--- Fp, i j; i, j = 1,2,.......12The reasoning is always the same; ifan event is low on F; then it will have tobe high on some Fj to make news at all.For a low Fi the probability that any F,is high is greater than fora high F. - since

The additivity hypothesis focuses ontype 1and the complementarity hypothesison type 2 - one might then add theobvious exclusionhypothesis hat would focuson type 3.We then turn to the presentation of ourdata and to a systematic testing and dis-cussion of a selection of the hypothesesmentioned.

3. The dataWe have selected four Norwegian news-papers, three international crises, and foreach crisis a number of variables to usein the content analysis of what the news-papers wrote about the crises. The ration-ales behind our selections are as follows.a. The newspapersNewspapers play an important role inNorwayl8 and appear in a relativelydecentralized pattern. However, when itcomes to foreign news only newspapers inthe bigger cities would give sufficientcoverage to merit a content analysis, andparticularly the newspapers in the capital,Oslo. They are 10 in number, and we haveselected 4 according to the followingdesign:


11/29

73Table 3. The newspapers in the sample, and their average circulation

Morning AfternoonRadical I II1960 67,494 1964 67,000 1960 98,352 1964 95,000Conservative III IV1960 21,204 1964 16,800 1960 37,040 1964 38,000

No. I is the government paper, no. IIis the afternoon paper of the widest circu-lation, no. III a conservative paper andno. IV is a conservative afternoon paper,of tabloid format. Apart from the conserv-ative morning paper, circulation figureshave been constant during the period. Asusual the morning papers are consideredmore 'serious', and indeed are, at least interms of lay-out, use of headlines and vo-cabulary - this holds true for the radicalas well as for the conservative press. Butthese political terms are generic terms anddo not imply that the papers in the samecategory follow the same party line.19The reason for this particular choice liesin the experimental design we obtain: bymeans of the two axes in the Table we geta sufficient dispersion to provide a settingforreplication. A finding that holds true forall four papers will receive a higher degreeof confirmation when the papers are dif-ferent than when they are similar.20b. The crisesWe have selected three crises, and ac-cording to a very simple experimental de-sign. We wanted both synchronic anddiachronic comparisons to check for con-sistency in the way news was presented inthe four Norwegian papers. For that rea-son two crises were selected that occurredsimultaneouslyand otherwise were roughlycomparable: the Congo and Cuba crises inthe summer of 1960. In 1964 a thirdcrisisoccurred that had some of the same char-acteristics, viz. the Cyprus crisis, and wedecided to give it the same kind of ana-lytical treatment. Obviously none of thethree criseshavevery definite pointsof initi-

ation and termination, so we had to selectmore or less arbitrary cutting points. Forthe Congo and Cuba these cuts were madeso as to coincide almost with the month ofJuly 1960, which will be remembered as arather conflict-laden one. For Cyprus themonth of March and the first half of April1964 were included. In three appendiceswe have given chronological surveys ofwhat happened in the three areas duringthe periods mentioned, according toKeesing'sContemporaryrchives.We do notclaim that they represent well-definedchapters in the books about these crises -nor is that in any sense essential for ourpurpose.It may be objected that these threecrises are much too special to give a basisfor assessing the structure of foreign news,and we would agree with that assertion.On the other hand, the three crises con-tain elements of particular interest andrelevance in the current world situation.The conflicts are acted out in theatersremote from the elite northwestern cor-ners of the world - but with traditionalpowers in that corner deeply involved -Belgium, France, the USA, Britain. Inall three cases world conflicts, both of theEast-West and the North-South variety,are superimposed on local conflicts or viceversa: local conflicts develop from worldconflicts. The UN intervened in the Congoand Cyprus conflicts and not in the Cubaconflict. In short, many elements of thecontemporary international situation arepresent. In addition the conflicts are sosimilar that roughly the same analyticalscheme can be used so as to obtain com-parability.


12/29

74c. The variablesThe 'unit of analysis' is the press cuttingas defined by the newspaper itself when ittypographically sets a unit apart from itssurroundings, such as a news story, aneditorial, an article (reportage, interview)or a letter to the editor - to quote thecategories we have used. The contextualunit is the newspaper itself, which meansthat for all units we have two kinds ofvariables: contextual variables referring tothe newspaper (its name, party color,

or Turkish) on the one hand and Belgium/France-US-Britain/Greece/Turkey on theother. The terms denote a difference inrank and a dependence relationship thatis basic to the crises - and nothing else.As can be seen at a glance the list coverssome but certainly not all of the factors weare interested in, according to the theory- nevertheless they can be brought to bearon a number of the hypotheses.Let us then give in Table 4 a brief surveyof the nature of our data:

Table 4. Surveyof thepress cuttingscodedCongo Cuba Cyprus

Paper News Other (N) News Other (N) News Other (N)I......... 89% 11% (112) 98% 2% (37) 96% 4% (171)II ......... 96% 4% (100) 89% 11% (43) 88% 12% (178)III ......... 90% 10% (103) 88% 12% (65) 91% 9% (241)IV ......... 87% 13% (108) 93% 7% (29) 92% 9% (75)Total....... 91% 9% (423) 91% 9% (174) 91% 9% (665)

date, etc.), and proper variables that referto the unit itself. These variables mayagain be subdivided into variables thatapply to the cutting as a whole (its pres-entation in the newspaper, its length, its'type' as above, the source in terms ofpress agency, person quoted, etc.) andvariables that apply to what is written.As to the latter we have worked with alist of items and coded a unit according tothe presence or absence of these items.More specifically we have been inter-ested in:

top leaderEast- East-West colonynegativeeconomic

Although it makes little sense to poolthe data from the four newspapers - thesample is made for replication and not forestimation - unless they can be shown tobe sufficiently homogeneous, it is interest-ing to see that 91% of the cuttings arenews messages in all three cases. Thirty-six units were editorials (divided 23-4-9on the three crises) and there were 58articles, etc. (divided 25-8-25). But outof a total of 1,262 pieces this means 3%and 5% respectively - and the numberof letters to the editor was less than 1% of

Underdog(colony)eliteWest-colony

positivepolitical

topdog(motherland)rank-and-file no people

motherland- UN- intra-colony colony colonyneutralsocial

(both or neither)cultural

We have used the generic terms 'under-dog-colony' and 'topdog-motherland' for,respectively, Congo-Cuba-Cyprus (Greek

the total. Although we shall certainly notneglect this 9% in the total picture wenevertheless feel justified in focusing the

1. Nation ......2. People ......3. Perspective....4. Mode.......5. Focus .......


13/29

analysis of what was written in these fourpapers on the news items.And this brings us to our main justifi-cation in presenting these data at all, theanswer to the obvious 'Who cares aboutfour Norwegian papers in the world? -at most half a million Norwegians'. We areconcerned with who has transmitted thenews and for each unit we have put downthe press agency or agencies quoted as asource:Table 5. The distributionf thenewsmessages nthepress agencies*

Congo Cuba CyprusAssociated Press(AP) 12% 20% 9%United Press Inter-national (UPI) 20% 17% 17%Reuters 31% 22% 52%Agence France-Presse (AFP) 19% 15% 24%Norsk Telegram-byra (NTB)21 28% 20% 54%Tass 1% 2% 0%Others 4% 1% 4%(N) (382) (158) (612)Own correspondent 2% 1% 7%*** Percentages do not add up to 100% sinceitems from different agencies are often pooledtogether in one unit, and sometimes there isno reference to agency.** 43 cuttings out of which 37 or 86% aredue to paper II - it had a correspondent inCyprus and this accounts for 23% of itscoverage.

The Table substantiates what we areafter: the sources of the news in theseNorwegian newspapers - among themost important politically - are inter-national. 'Own correspondent' is quiteinsignificant as a category, which is notstrange, taking into consideration the

75economy of Norwegian newspapers andthe distance to these particular theaters.This means, briefly, that foreign pressagencies account for about 95% of thenews items about these three crises, andthe news items account for 91% of thetotal number of units appearing in thesepapers about these crises. That, in turn,means that four foreign press agenciestake care of 95% of 91% or about 87%/,of the total - which again means thatNorwegians, in casu, are rather dependentfor their images of how the internationalsystem functions on the quality andquantity of news delivered through theseagencies.22

Consequently our study is a study of apart of the foreign news system, using howit impinges on four Norwegian news-papers in three crises as twelve casestudies.4. Testing the hypothesesWe then proceed to tests of the theoryof the structure of foreign news with thedata we have. All we can do is to exhaustthe possibilities our data give for tests ofthe theory, and we start with:H1: The mnoreistant the nation, the higher the

tendencyto reportelite action.If a nation is 'distant', here of coursetaken in the cultural and not in the geo-

graphical sense, identification with rank-and-file people will be correspondinglylow. At the same time, to become newsevents will have to fulfill some other re-quirements. We shall first test this hypo-thesis on the Congo and Cuba crises sincehere what is close and what is distant is soobvious. The data are as in Table 6:

Table 6. Distance of nation and rank of people reportedCrisis Congo CubaNation Congo Belgium % diff. Cuba USA % diff.Top leader ...........4...... 47% 15% +32 53% 23% +30Elite ....................... 50% 40% +10 47% 39% t- 8Common people ............. 30% 35% - 5 24% 6% +18(N) ........................(158) (382)


14/29

76Reading downwards one would expectdecreasing percentage differences accord-ing to the hypothesis; each percentagedifference should be lower than the pre-ceding ones. This holds in five out of thesix cases and gives us a degree of confir-mation of 0.84 using that as a criterion.Rank-and-file Americans went under-

Turks and Turkish Cypriots are hardlyvery different relative to, say, Norwegians,although the Greeks are probably some-what more familiar. But Britain can beset apart relative to the other four.The data, in Table 7, are quite interest-ing and confirm more or less everythingwe have said, with one important proviso:

Table 7. Distanceof groupand rank of peoplereported,CyprusGreek TurkishCypriots Cypriots Greeks Turks British

Top leader ................... 30% 7% 11% 13% 5%Elite ......................... 4% 4% 6% 5% 12%Common people ............... 40% 34% 6% 9% 15%(N)..........................(612)reported in Cuba - and this is certainlyunderstandable if one compares this withthe Congo case. There is a difference be-tween nationalizing industrial enterprisesand open conflict: the former is moreabstract and does not give rise to the sameamount of stories about women and chil-dren, for instance.Let us then turn to the data from Cy-prus. They are complicated by two factors.First of all, we do not at all have a clearpattern with two parties to the conflict, asbetween the Congo and Belgium andbetween Cuba and the USA in July 1960.At least five parties are involved at threelevels of dominance. At the bottom thereis the conflict between Greek CypriotsandTurkish Cypriots. Related to either andat an intermediate level is the conflictbetween Greece and Turkey, partly overCyprus, partly over anything else thatcan be added. Above that level againlooms Britain and other big powers, butwe have concentrated on Britain becauseof her past as a colonial power in Cyprusand her present as the holder of majorair-force bases on the island. Thus, thereis no clear bipolar pattern here as in theother two cases, with clear foci of identi-fication.23

Secondly, the ethnical dimensionis hard-er to apply. Greeks and Greek Cypriots,

The British are the only case where thereis a clear increase in the percentage ofreporting as we move down the socialladder. This is in agreement with thehypothesis of closest identification withthe UK since none of the other four per-centage-sets have a correspondingly clearand monotone pattern. But if we look atthe rest an interesting suspicion emerges:that there has been more identificationwith the Turks than with the Greeks. Al-though the differences are small they arethere: the common people/top leader ratiois 1.3 for the Greek Cypriots and 4.9 forthe Turkish Cypriots, 0.6 for the Greeksand 0.8 for the Turks. Of course, the over-reporting of Makarios may be explainedon such bases as the fact that he was wellknown from the fight against the Britishand that he, not the Turkish minorityleader Dr. Kutchuk is the President ofCyprus. Nevertheless, there is the suspi-cion of differential identification builtinto the structure of the news. This is nota part of the hypothesis, however, sincecultural distance can hardly be said to bea factor here. But it is nevertheless con-sistent with what was reported by oneNorwegian with special insight into thearea, as to British and Turkish views andperspectives being favored in general inthe Norwegian press.24


15/29

A particular way of looking at this hypo-thesis is by counting the number of timespeople are quoted in the news. To manyreaders it will be obvious that commonpeople are quoted in only 2 of the 612cuttings from the Cyprus crisis, to takeone example (more frequently in theCongo news where relative identificationwith the Belgians was stronger). The eliteand the top leaders are very frequentlyquoted and in general the ratio betweencuttings where people are mentioned withquotes and mentioned without quotes de-creases rapidly with decreasing rank. This,then, is one more factor of identificationthat favors the elite.

These general findings were checkedfor newspaper and stood up against thetest - the finding is replicated.After the relationship between distanceof nation and rank of the person reportedwe turn to the relationship between whatwe have called 'mode' of reporting andrank of persons reported. We have codedfor each cutting whether it reports some-thing 'negative' (something is destroyed,disrupted, torn down) or something 'posi-tive' (something is built up, constructed,

77be deduced from the complementarityprinciple isH2: The lower the rank of theperson, the morenegative the event.

We tested this hypothesis not onlywithin each crisis but also for each countryor nationality, which gives a total of ninecases (Cuba, the US, the Congo, Belgium,Greek Cypriot, Turkish Cypriot, Greece,Turkey, Britain). For each case we hadthree ranks (top leader, elite, rank-and-file) and for each rank the event could beeither negative or positive since we dis-carded all cuttings that were not clear.To arrive at a composite measure - 1 wasgiven to a negative cutting and +1 to apositive one, and the average 'mode' wascomputed. Thus, with 80 negative cuttingsand 35 positive we would get:-80 + 35 -45__80 + 35 115In general the index ranges from - 1 to+1, but the limits were hardly attained;the news was almost never completely one-colored.

The data are as in Table 8:Table 8. Rank of thepersonand modeof the event

Top leader Elite people Common people1. Cuba ...................... -0.66 -0.51 -0.592. USA ....................... -0.91 -0.72-1.03. Congo ...................-... 0.39- 0.45 - 0.474. Belgium..................... -0.08 -0.33 -0.455. Greek Cypriots ............. + 0.16 -0.20 -0.386. Turkish Cypriots .......... + 0.13- 0.43- 0.477. Greece ..................... + 0.21 -0.10 -0.208. Turkey ..................... - 0.26- 0.18- 0.389. Britain ..................... + 0.39 - 0.20- 0.33

put together) and we are interested inknowing, for each of our three rank levelswhere people are concerned, whether theyare seen in a context that is negative orpositive (we omit the cuttings where theevent-context is coded as 'neutral', 'both'or 'neither'). The obvious hypothesis to

Out of the nine cases the hypothesis isclearly confirmed in cases nos. 3, 4, 5, 6,7 and 9; which means in 6 of the 9 cases.In the other three the trend from 'elite' to'common people' is as it should be, butthe three top leaders are seen in a morenegative context than one would predict


16/29

78from the theory. Two of these cases, nos.1 and 2, concern Cuba-US relationswhich were then in a very critical phasewith both top leaders declaring negativeactions against one another in an escalat-ing sequence. The third case has to dowith the Turkish top leader who in thatperiod used a language characterized bythreats and invasion menaces. (See Ap-pendix III.) However, any exceptionfrom the rule can always be 'explained'away by invoking some special circum-stances, so we shall be satisfied by noticingthat out of a total of 27 numerical relation-ships (three for each case) 23 or 85% arein the expected direction, i.e. increasinglynegative with decreasing rank. This is

them as causes - is higher, due to the eliteconcentration reported above. But theproportion of negative events relative tothe total number of contexts is higher thelower down in society the news comesfrom. Or, in other words, whefe positiveevents are reported they will be morelikely to occur as contexts for an eliteperson than as something surrounding thecommon man.

However, one thing is context, anotherthing is who is seen as the cause of theevent. According to our hypothesis onewould expect the same pattern if the newsstories are analyzed for their tendency toattribute causes to somebody, and this iswhat we actually get, as Table 9 shows:

Table 9. Rank of causalagent of an eventand the modeof the event*% negative inYule Percentage the news fromcorrelation, Q difference the group

1. Cuba........................ 1.0 21% 79%2. USA .......... 1.0 4% 96%3. Congo ....................... 0.47 15% 91%4. Belgium...................... 0.46 18% 76%5. Greek Cypriots................ 0.80 44% 65%6. Turkish Cypriots .............. 1.0 25% 94%7. Cypriots, not specified ......... 0.93 51% 87%8. Greece ...................... 0.76 40% 86%9. Turkey ...................... 0.57 18% 90%10. Britain ...................... 0.57 27% 67%

* Top leader and elite have been pooled together.high even though the 27 relationships arenot independent.The general finding was checked fornewspaper and for press agency and stoodup well against that additional source ofvariation.

Thus, we have been brought a stepforward towards the idea that commonpeople have to appear or be presentedin a negative context to make news, muchmore than is the case for people higher up- relatively speaking. It may well bethat the total volume of negative eventsreported in connection with elite people- whether happening to them or having

Thus, the hypothesis is confirmed in 10out of 10 cases: the lower the rank of thecausal agent, the lower also the chancethat he is seen in a context of somethingpositive happening.From here we may turn to the nexthypothesis, thus completing the trianglewe have made of variables:H3: The more distant the nation, the morenegative the event.

The data in Table 9 are relevant for thehypothesis and do not appear to give anyclear pattern of confirmation. We haveused the data about causal agents only,


17/29

not the data where a nation may alsoappear as the victim of a negative act.As can be seen the British, as a total, withthe three rank categories pooled together,appear as the causes of negative events lessfrequently than do the others, with theexception of the Greek Cypriots. Corre-spondingly, the Belgians are causes ofnegative events less than the Congolese.But the US are producers of negative actsmore than the Cubans. One reason maybe that these negative acts were actuallynot seen as negative by the agencies re-porting them, often AP and UPI - butas adequate reactions in a situation ofintense conflict.

It should be kept in mind that although

79Another way of looking at these data nowis to ask the question: Cuba, the Congoand Cyprus are far away places, they are'culturally distant' (factor F4.1). How do

the events come to be represented at allas news ? Because they are made 'relevant'(factor F4.2). Thus we get the hypothesisH4: The more culturally distant the theater,the more relevant must the event appearto be.

Unfortunately, we do not have datafrom theaters with a wide range in culturaldistance, but we can get some idea aboutthe validity of this hypothesis in Table 10from the distribution of the news storieson what might be called the 'perspective':

Table 10. Location f the theaterandperspectivef thenewsstoryCrisis RelationsbetweenEast- East- West- Motherland- UN- Intra- (N)West 'colony' 'colony' 'colony' 'colony' 'colony'Cuba......... 9% 35% 22% 59% 9% 20% (158)Congo......... 9% 20% 20% 52% 41% 36% (382)Cyprus........ 0% 4% 8% 54% 59% 29% (612)the material includes all news from theCongo, Cuba and Cyprus in the periodsof analysis, this is not the case with the'mother countries': Belgium, the USA, theUK, Greece, and Turkey. News from thoseplaces which have no relevance to theareas of crisis have appeared in the papersin addition to the coded items. (Just tomention one example: the death andfuneral of King Paul of Greece took placein the middle of March, causing Greeceto appear quite heavily in the news storiesfor a week.) In the case of culturallyclose countries like the UK, the USA andBelgium, a great variety of news storiesreaches the papers every day, regardlessof major events. This would serve tobalance off the negative impression thesecountries give as partners in the colonialcrises analyzed. Nevertheless we do notfeel that H3 has been confirmed, althoughit has not been disconfirmed either.

A purely cold-war perspective involvingEast and West alone has not been mademuch use of, but East and West reappearin their relations with the 'colony', thusincreasing relevance by linking the con-flict to the East-West system. Most im-portant, of course, is the 'motherland-colony' perspective, appearing in morethan half of all the news stories. It may besaid that this was what the conflicts wereabout, but that is not so obvious. Thereare many ways of presenting an event, andparticularly many ways of presenting whatto many appeared as fights of independ-ence. Thus, nationalization of industries,independence of a new nation and thefight between a majority and a minoritymight all have been presented as fairlyinternal events with local actors only andthe 'motherland' appearing more as aconstant condition that could be men-tioned in, perhaps, 10% of the stories.


18/29

80But this would have presupposed a muchhigher degree of identification, up to thelevel one probably had in the newspapersfrom the Congo, Cuba and Cyprus duringthose periods. Instead, events are seen asunfolding themselves in the periphery ofthe 'motherland' with no real local auto-

ain in 31% of the stories, which is a highpercentage if one conceives of the conflictas essentially limited to the Greece-Turkey-Cyprus triangle.

There is another way of testing thehypothesis, taking as in Tablel 1 the focusof the news story as point of departure.Table 11. Locationof the theaterandfocus of the news story

Crisis FocusEconomic Political Social Cultural (N)Cuba ............... 63% 56% 0% 1% (158)Congo............... 13% 86% 11% 2% (382)Cyprus.............. 10% 95% 1% 0% (612)nomy. The 'colonies' are not causallyself-sufficient. News stories that have an'intra-colony' perspective exist, but thereare two simple explanations for that. Thefirst one is in terms of F12- the idea thatevents will have to be negative - andsimply refers to the fact that in both theCongo and Cyprus local conflict is at theroot of the 'crisis'. Even under this con-dition, however, the local conflict is notenough: some familiar groupings, suchas the East, the West, the motherland, theUN, have to be added to it to make itreally newsworthy. And the other explana-tion is in terms of F7: since the theater isalready in the news it will probably re-main in the news because an apparatushas been established that requires a certainquantity of news stories to be maintained.That there are explanations of themechanismunderlying this in terms of suchfactors as the nationality of the pressagencies and the training of some of thejournalists, whether they are foreign orlocal by nationality, is obvious. Thus, onewould expect the news from Cyprus tobe much concerned with events seen asrelevant to Britain, since the British agencyReuters appears in connection with 52%of the news stories. One consequence ofthis is found in the circumstance thatCyprus is seen in its relationship to Brit-

The figures are quite clear and tell thekind of story one would predict: the threecountries in the world periphery enter byway of variables that link them to thecenter part of the world. 'Social' and 'cul-tural' are more internal and do not havethe same immediate ramifications to thetopdog nations. The only variation in thefocus is from economic to political as onemoves from Cuba via the Congo situationto Cyprus, and that corresponds well tomost images of 'what really happened'.

Again, the findings hold up against thevariation in newspaper and press agency.We let this suffice as an indication of howhypotheses derived from the complemen-tarity theory can be tested.It may be objected that what we havesaid is an artefact of the three crises wehave picked for our sample. There is noother way of exploring this objection thanby means of a new project.5. Discussion

On the basis of what we have presentedwe feel that it would be unreasonable notto have some confidence in the generalhypothesis. There is probably such a phe-nomenon as complementarityof news factorsalthough much remains to be done interms of refinement of the hypothesis.Under what conditions will the effect be


19/29

more pronounced or less pronounced,which pairs of factors tend to produce thestrongest (or the weakest) effect of com-plementarity, and how do the factors com-bine in patterns of three and four, etc.?We leave this for future research, andturn to a discussion of what this implies- under the assumption of the generalvalidity of our thesis.A discussion of this kind can best bedone by selecting out of the 66 possiblepairs some crucial pairs that are particu-larly important in terms of their conse-quences for the kind of image of the worldthat they will promote. The numbers referto the list of twelve factors. These hypo-theses are bivariate only, and comingresearch in this field will have to carry thethinking and the analysis up to, at least,the level of three variables or factors atthe same time.

(1,4): The more distant a nation is, themore will an event have to satisfythefrequency criterion.In other words: the distant na-tion will have to produce eventsthat capture attention particu-larly easily in order to be recor-ded. The consequence of this isan abruptnessand unconnected-ness that the news from suchcountries will display. Naturaldisastersand accidents will playan important role, and changesof government. The build-up ofevents, based on small quanti-tative changes, will go under-reported - it is only when theylead to the big qualitative chan-ges that they make news. Thisagain may provide readers withan image of these countries asplaces where things happen allof a sudden and in an unpre-dictable way - in other wordsinherently dangerous and in-herently different places.

(4,3): The more distant an event, the lessambiguouswill it have to be.The remote and the strangewillat least have to be simple if itis to make news - complexitiescan be taken care of if theyare found within one's own cul-

81ture, but not if they are foundat a considerable distance. Theimplication of that is obvious!The culturally distant acquiresa presentation in terms of 'idealtypes'; whole nations and con-tinents are described in sweep-ing terms and this may leavethe impression of a uniformityand homogeneity that is notpresent in the reality of thatnation. One's own nation isdescribed in complex termswhich will correspond more tothe idea most people have of a'civilized existence'. On theother hand, the phenomenonindicated here will tend tofoster the idea of a simple, pri-mitive and more 'human' kindof existence in remote countries.

(4,5): The moredistant thenation, the moreconsonantwill the news have to be.For a far away nation to makenews it will be particularly ne-cessary that the news should fita pattern of expectation. Thusthe golpe militar in Latin Ameri-ca, according to this hypothesis,will make news exactly becausethat is what is expected - it is acase of news being 'olds'. Theopposite development will notso easily fit the expectation pat-tern and for that reason lesseasily be reported, becauseprobably it will, consciouslyor unconsciously, be registeredas a quasi-event that will notlast. Any story of sexual extrem-ism from Sweden will receivean a priori credibility that willmake it pass many filters ofnews communication, whereasstories or statistics to the oppo-site effect may be seen as atypi-cal or even fake and propa-ganda, and not be reported.The consequence of this will bethat distant nations appear asessentially unchangeable where-as one's own cultural sphereundergoes real, basic change.

(7,2): The higher the continuity effect, thelower can the thresholdbe.We only list this hypothesis forthe sake of completeness, for itis actually the definition of the


20/29

82continuity effect. Once an eventhas 'made it' the news channelwill be more readily open forthe follow-up events, at a lowerthreshold value. The effect ofthis will be the creation of'news strings' that may createartificial continuities just be-cause the channel is open.

(8,2): The higher the composition effect,the lower can the thresholdbe.This is also listed for the sakeof completeness, since it is al-ready included in the definitionof the composition effect. Theidea is simply that news can en-ter because of under-represen-tation of categories that shouldbe represented according tosome overall judgement - notbecause they are important bythemselves. This, in turn, meansthat in periods where little elsehappens abroad the limit defin-ing newsworthiness may bedrastically lowered so as to in-clude news items that score rela-tively low, and this in turn mayproduce images of discontinuitythat do not correspond to thereal world.

(1,12): The less negativethe news, the moreimportantthefrequency condition.This is already referred to inconnection with the theory forthe negativism of the news. Buthere it is put in a stronger form:positive events will have to beparticularly short of durationto appear as news. This means,essentially, that a premium willbe put on the ability to makeceremonies where developmentscan be telescoped into an eventthat is reportable. Obviouslythe more elite people can beadded to it the better for thenewsworthiness, and this has adouble effect. First of all it maycontribute to a false image ofhow positive developments comeabout, since the amount ofplanning and painstaking work,mostly and in most cases bynon-elite people, goes under-reported. Secondly it forcesmany people into a kind of ac-tivity usually referred to as P.R.

- public relations - that isoften accepted as a part of theirwork, where one might questionthe wisdom of the structureof the news communicationinstead.

(9,4): The lower the rank of the nation, thelower must the cultural distance be.This only means that if a nationis low in terms of rank it mustcompensate for that in termsof proximity. Or in other words:the topdog nations of the worldwill each have their own set ofunderlings that they over-reportfrom, relative to what they re-port from other low rank na-tions. For the U.S. it will beLatin America, for France Com-munaute Francaise, for Britainthe Commonwealth countries,for the Soviet Union the socialistbloc, for China (probably) se-lected countries in South andFar East Asia. This pattern,in turn, will tend to reinforceexisting divisions of the worldsince reporting will probablymake for some kind of identifi-cation.

(9,5): The lower the rank of the nation,the more consonant will the newshave to be.This is very similar to the hypo-thesis of the relationship be-tween distance and consonance- but whereas that hypothesisemphasized consonance withwhat one would expect frommore or less stereotyped con-ceptions of a foreign culture, weare here concerned with stereo-types about low rank. The typi-cal example would be news thatemphasizes the difficulties lowrank nations have: signs of'immaturity' in terms of pay-ment crises, political instability,murder at the top of society,etc.. The consequences are thesame as for hypothesis (4,5)above.(10,6): The lower the rank of the person,the more unexpectedwill the newshave to be.This has actually been touchedupon in different contexts al-


21/29

ready and the idea is simplythat whereas elite people canhave their day-to-day routinereported, rank-and-file peoplewill only make news whensomething happens that standsin a very marked contrast totheir ordinary existence. Thegood examples are sudden ac-quisitions of wealth and nega-tive actions.(9,10): The lowerthe rankof thefnation,thehigherwill a personhaveto beplaced n thatnatzono makenews.This may lead to an image ofthe world underdog nations as

extremely elite-dominated witha non-existingmassof rank-and-file people. In political termsthis image will probably tendto reinforce the conditions thatmake such images warranted.This will also make for pooridentification particularly ifelite action in low rank nationsis also negative.(9,12): The lower the rankof the nation,themorenegativewill thenews rom

thatnationhave obe.In other words,when somethingpositive and good is happeningit will have to be located in ahigh rankingnation - from theunderdog nations of the world,typically, news reports will beoverwhelmingly negative. TheLatin American proverbial caseof the galpe militar is one ex-ample; all the disasternews fromsuch nations is another. Thethesis is that positive things thathappen in the underdog coun-tries will go under-reported andthis will promote an image ofthose countries as being unableto govern themselves, and asinherently inferior to the top-dog countries.(10,12): The lower he rankof a person, hemorenegativewill his actionshaveto be.In other words, the thesis is

that common people must dosomething negative to makenews, and the lower down theperson is, the more negativeshould it be. At the bottom of

83society one enters the newspages more easily as a criminal- but sport should of course bementioned as the big compen-sating mechanism. It may alsobe regarded as so important,together with the arts and en-tertainment, that it actuallyinvalidates the hypothesis. Nev-ertheless, the kind of positiveaction the rank-and-file personhas it in his power to performis, perhaps, more likely to beof a kind that will never makethe news - not only for thereason mentioned in the hypo-thesis but because it does notsatisfy the criteria of frequency,threshold, unexpectedness andcontinuity either. If the ordinaryman is to enter positively, itwill probably have to be in anarticle, reportage, etc. It maybe objected that he enters whenhe wins in the lottery - butthis is not an act of his - ithappens to him, like a cata-strophe. The implication of allthis may easily be a kind of re-inforcement of class society inthe sense that the top is over-represented with the good andthe positive that occurs, andthe lower layers of society areportrayed as producers of lessfortunate events.

(11,12): The less personal the news, themorenegativewill it haveto be.In other words,when somethingpositive happens it is morelikely to be attributed to people,whereas something attributedto non-people will have to benegative to hit the news. In asense this may also be seen as areflection of the dominant ideaof man as the maker of his ownprogress against the forces ofnature that tend to inundatehim with floods, shake him topieces with earthquakes, etc.It may be worth while to collect to-

gether what has been said about nationsthat are culturally distant and nationsthat are low in international rank.25Wecan combine it because what we have said


22/29

84should a fortiori be valid when these twocriteria - negative for newsworthiness -are superimposedon each other. In short,from such countries news will have to referto people, preferably top elite, and bepreferably negative and unexpected butnevertheless according to a pattern that isconsonant with the 'mental pre-image'. Itwill have to be simple and it should, ifpossible, provide the reader with somekind of identification - it should refer tohim or his nation or group of nations. Thiswill, in turn, facilitate an image of thesecountries as dangerous, ruled by capri-cious elites, as unchanging in their basiccharacteristics, as existing for the benefitof the topdog nations, and in terms oftheir links to those nations. Events occur,they are sudden, like flashes of lightning,with no build-up and with no let-downafter their occurrence - they just occurand more often than not as a part of themachinations of the ruling or oppositionelites.The consequence of all this is an imageof the world that gives little autonomy tothe periphery but sees it as mainly existingfor the sake of the center - for good orfor bad - as a real periphery to the centerof the world. This may also tend to amplifymore than at times might seem justifiedthe image of the world's relatedness.Everything's relevance for everything else,particularly for us, is overplayed. Itsrelevance to itself disappears:

Mr. Mboya complained of thePress (foreign-owned) in Africa be-having and writing as though it wereoperating in London, Paris or NewYork 'where the problems and anxi-eties are entirely different from thosecurrent in Africa'. He said these andmany other questions kept comingup in the minds of many Africans asthey try to figure out what freedom ofthe press meant in the African con-text. He was of the view, therefore,that it was important that the Pressshould concern itself with finding outwhat goes on in the African mind.The world's verdict on Africa, how-

ever, was often reduced from sub-jective dispatches of foreign journal-ists paying short visits to the variousparts of Africa. The result was thatnews coming out of Africa was oftenrelated to the already biased and pre-judiced mind that keeps asking suchquestions as: 'is this pro-East or pro-West?' but nobody asked: 'is thispro-African ?'26This is particularly dramatic in connec-tion with new countries. Their newness,which is probably the major fact for themajorityof their inhabitants, is not stressedexcept as reports from the independence

ceremony if there is any (because it satis-fies F1). Instead the news is interpreted ina context of the old, and since all threecountries were centers of major events inthe periods we have analyzed, they haveprobably for many people come to bedefined for some time through these crises.This, in turn, may influence people's be-havior towards the nations in question,and if they are very young natiornsserveas a kind of imprinting experience,27withthe consequences that implies for laterrelationships. It would be interesting toknow something more precise about howfar behind political independence whatone might call causal independence (orauto-causation, causal autonomy) is lag-ging.As to the developed countries the gene-ral implication of what we have said hasalready been mentioned: conflict will beemphasized, conciliation not.286. Some policy implicationsThe policy implications of this articleare rather obvious: try and counteract alltwelve factors. More specifically, thismeans:

1. More emphasis on build-up andbackground material in the total mediaoutput. Journalists should be bettertrained to capture and report on long-term development, and concentrate lesson 'events'.


23/29

2. Occasional reports on the trivialeven if it does not make 'news', to counter-balance the image of the world as com-posed of strings of dramatic events.

3. More emphasis on complex and am-biguous events, not necessarily with anyeffort to interpret them.4. More reports from culturally distantzones even if the content has no immediaterelevance for oneself. Experiments withnewspapers in different countries exchang-ing local columns might prove even moreinteresting than reprinting what was saidin the newspaper fifty or a hundred yearsago.5. More emphasis on the dissonant, onthat which does not fit stereotypes. Train-ing of journalists to increase their insightsinto their own stereotypes so as to facilitatetheir awareness of the consonance factor.6. More emphasis given to the predict-able and frequent, for the same reason asunder 2 above.

7. More awareness of the continuityfactor - and at the same time more em-phasis on follow-ups even if the chain ofevents has been interrupted for some time.Often one has the impression that onehears about something negative that hashappened but not about how it has beencounteracted, if the time-span is so longthat the continuity has been broken.

858. More awareness of the compositionfactor in order not to create news arte-facts.9. More coverage of non-elite nations.10. More coverage of non-elite people.11. More reference to non-personalcauses of events. Special training is prob-ably needed here.12. More reference to positive events.These implications work on one factorat a time and would, if implemented,reduce the effects of the factors. However,the combined effects of the factors might

still persist even if the effect of any onefactor is reduced.One might say that all or much of thisis what the elite paper tries to do, and thatis probably true. However, elite papersare probably mainly read by elite peopleand this may increase the distance be-tween center and periphery where inter-national perspective is concerned.Hence one additional need is for a morewidely dispersed style of news communi-cation in agreement with these principles.It should be emphasized, however, thatthe present article hypothesizes ratherthan demonstrates the presence of thesefactors, and hypothesizes rather than dem-onstrates that these factors, if present,have certain effects among the audience.

TE S* This is a much revised and extended version of a paper presented at the First Nordic Con-ference on Peace Research, Oslo, 4-8 January 1963 and as a guest lecture at Danmarks Journa-listhojskole, Arhus, May 1964, here published as PRIO publication no. 14-2. The authors wishto express their gratitude to the Institute for Social Research, the Norwegian Research Councilfor Science and the Humanities, and the Norwegian Council for Research on Conflict and Peacefor financial support; to stud.mag.art. Marit Halle and stud.mag.art. Elisabeth Bogh for assist-ance with the data-collection and to our friends and colleagues at PRIO and particularly toEinar Ostgaard for stimulating criticism and suggestions.1 Einar Ostgaard, 'Factors Influencing the Flow of News', Journal of Peace Research,1965,no. 1, pp. 39 ff.2 For an interesting article making systematic use of these two indicators of interdependence,see Kaare Svalastoga, 'Technology and Autonomy', Acta Sociologica,Vol. 5, pp. 91-99.3 Thus, a completely realistic image of other people's image of oneself might have a harmfuleffect on the social adequacy of one's behavior. Thus, there is the important finding by Caplowand McGee (The AcademicMarketplace, ew York: Basic Books, 1958) that members of organiza-tions are often subject to an Aggrandizement Effect whereby they overestimate their own organ-


24/29

86ization relative to others in the field. One might argue that if they did not, the consequentdrop in self-image would result in lower achievement levels. And this may have a parallel in thefield of international affairs: if the news structure was symmetric, giving to each nation its due,relative to how it was estimated by other nations, an important source of self-pride and asser-tiveness might be too weak to spur effective action.4 For one way of describing this chain see Johan Galtung and Mari Holmboe Ruge, Pre-sentasjonen v utenriksnyheterOslo: PRIO stencil no. 14-1, 1962), pp. 71-78.

5 Ostgaard, op. cit., pp. 42 f.6 For an impression of what sociologists can get out of the condition of sleeping see VilhelmAubert and Harrison White, 'Sleep: A Sociological Interpretation', ActaSociologica,Vol. 4, No.2, pp. 46-54 and Vol. 4, No. 3, pp. 1-16.7 This, of course, is a fundamental idea in the psychology of perception. Actually there aretwo separate ideas inherent here: the notion of an absolute level that must not be too low, andthe notion of the increase needed to be noticed - the 'just noticeable differences' (jnd's). Thejnd increases with increasing absolute level; the stronger the amplitude, the more difference isneeded to be noticed (whether this is according to Weber's principle or not). This principleprobably applies very explicitly to news communication: the more dramatic the news, the more

is needed to add to the drama. This may lead to important distortions. The more drama therealready is, the more will the news media have to exaggerate to capture new interest, which leadsto the hypothesis that there is more exaggeration the more dramatic the event - i.e. the lessnecessary one might feel it is to exaggerate.8 N. R. Ashby in An Introductiono CyberneticsNew York: Wiley, 1957) defines noise simplyas distortion that may create differences in interpretation at the sender and receiver ends of acommunication channel. But one may just as well say that the signal distorts the noise as viceversa.9 B. Berelson and G. A. Steiner in their HumanBehavior!An Inventoryf ScientificFindings(NewYork: Harcourt, Brace & World, 1963) mention a number of principles under 'Perceiving', andtwo of them are (p. 112 and p. 100):B7: The greater the ambiguity of the stimulus, the more room and need for interpretation.B3.3a: There may also be decreased awareness of stimuli if it is important not to see (per-

ceptual defense).What we have been doing is to combine these theorems (but not deductively) into the idea ofdefense against ambiguity. There are several reasons for this. Modern newspapers are mass mediaof communication, at least most of them, and publishers may feel (justifiably or not) that increasein ambiguity may decrease the sales. Moreover, to the extent that news shall serve as a basis foraction orientation ambiguity will increase rather than reduce the uncertainty and provide a poorerbasis for action.10 The common factor behind both dimensions of what we have called 'meaningfulness' isprobably 'identification'.11 Again, some findings from Berelson and Steiner are useful (op. cit., p. 101 and p. 529):B3.2: With regard to expectations, other things equal, people are more likely to attend toaspects of the environment they anticipate than to those they do not, and they are morelikely to anticipate things they are familiar with.B3.3: With regard to motives, not only do people look for things they need or want; butthe stronger the need, the greater the tendency to ignore irrelevant elements.Al: People tend to see and hear communications that are favorable to their predispositions;they are more likely to see and hear congenial communications than neutral or hostileones. And the more interested they are in the subject, the more likely is such selectiveattention.12 For a discussion of this see Johan Galtung, 'Summit Meetings and International Relations',Journal of Peace Research,1964, pp. 36-54.13 For a discussion of this factor see Ostgaard, op. cit., pp. 151.14 Festinger has a very interesting account of how Indians selected rumors following an earth-quake, and consistent with the fear provoked by the earthquake: 'Let us speculate about thecontent of the cognition of these persons. When the earthquake was over they had this strong,

persistent fear reaction but they could see nothing different around them, no destruction, nofurther threatening things. In short, a situation had been produced where dissonance existedbetween cognition correspondingto the fear they felt and the knowledge of what they saw aroundthem which, one might say, amounted to the cognition that there was nothing to be afraid of.The vast majority of the rumors which were widely circulated were rumors which, if believed,


25/29

87provided cognition consonant with being afraid. One might even call them "fear-provoking"rumors, although, if our interpretation is correct, they would more properly be called "fearjusti-fying" rumors.' Leon Festinger, 'The Motivating Effect of Cognitive Dissonance', in GardnerLindzey (Ed.), Assessmentf HumanMotives (New York: Grove Press, 1958), p. 72.

15 As an example some impressions can be given from three months systematic reading ofthe Moroccan newspaper Le Petit Marocain.In very summarized form: the first page containednews about progress in Morocco, the second about decadence, murder, rape and violence inFrance - so that anybody could draw his conclusion. Of course, such things will depend ratherheavily on the value-systems of the editorial staff - but we nevertheless postulate the existenceof general patterns. Ola MArtensson,in a mimeographed report (in Swedish) of a content ana-lysis of three major papers in the USSR, indicates both personification and elite concentration.Ola Martensson, Pravda,IzvestijaochKrasanajaZvezdaundervdren chhisten1964 (Lund: Institutefor Political Science, Lund University, Sweden, 1965), 26 pp. mimeo.16 Ostgaard, op. cit., pp. 52 ff.17 As an example it can be mentioned that in a survey carried out in Norway, November-December 1964, to the question 'What do you think has been the most important event in thenews recently?' 53% answered in terms of elections in the US and changes of power in the Soviet

Union, i.e. in terms of the top elite people in the top elite nations. The next answer category,'events in the Congo', made 9%.18 Norway appears as No. 7 in a list of 125 countries, according to the UN StatisticalYearbook1962 (New York: United Nations, 1963). The variable used is daily newspaper circulation per1,000 population.19 We omit the names of the papers, not so much out of considerateness, which would be outof order - firstly because we have nothing really inconsiderate to say about them, and secondlybecause they are public phenomena that might well be subject to public appraisals. The names,however, are of interest to Norwegian readers only and carry many connotations that will notbe used in the analysis anyhow.20 Not included in our sample is the biggest Norwegian newspaper with two daily issues andan average 1964 circulation of 168,000. This paper is the one with the most complete coverageof international events. Its political attitude is moderately conservative. The reason for excluding

this paper is not only the considerable extra work of coding that it would imply, etc., but alsothe fact that we were primarily interested in papers with a very clear political profile that wouldspan the political spectrum better.21 One thing that should be explained is 'Norsk Telegrambyra'. Only in some very few cases(nine in all) in connection with Norwegian soldiers recruited for UN service in the Congo doesthis mean that the Norwegian agency was actually the source of the piece of news. In all othercases NTB actually stands for AFP or Reuters and other Scandinavian agencies, since it acts asan agent for these foreign bureaux in Norway. (UPI and AP have their own offices in Oslo.)22 With the exception that the two less wealthy conservative papers have chosen one each ofthe American agencies (III does not subscribe to UPI and IV not to AP), all four papers weresubscribers to these agencies in 1960 and 1964. And all four agencies (AP, UPI, Reuters, AFP)have their headquarters in the three major powers of the world's Northwest - a region whereNorway is also located.23 It should be noted that the time-span covered in the Cyprus crisis is of a somewhat dif-ferent nature than for the Congo and Cuba. The building up of the Cyprus crisis actually took along time and only culminated during the winter months of 1963-64. For comparative reasonswe still decided to analyze a short period. We chose the weeks immediately prior to and duringthose in which the UN was actively brought into the conflict, in order to be able to make acomparison with the Congo situation. But this excluded the period in which Greek-Turkish-Cypriot relations were most strained, probably in January-February.24 Bjarte Kaldhol, 'Norske soldater til Kypros', Dagbladet,12/2 1964, p. 4.25 Wilbur Schramm refers in his book, Mass Mediaand NationalDevelopmentParis: UNESCO,1964, p. 64), to an investigation where Indians have commented on the way India is presented inthe American press. Four aspects of this particular news communication are resented, and it isclaimed that they represent the greater part of the news total: India in the East-West power

struggle and communism, American economic aid to India, stories about disasters and hunger,and stories about 'bizarre and outlandish things' in connection with child marriages, untoucha-bility, etc. The first two are typical examples of increased relevance by tying what happens inremote places to one's own country, the third is a clear case of F12,but also of a case where distantcountries are presented as victims rather than agents of what happens. The last factor is a case


26/29

88of F5.1 - it is consonant in the sense of being predicted from 'knowledge' of Indian culture;it fits stereotypes.26 The IndianExpress,11 July 1962, p. 6.

27 For some comments on the phenomenon of 'imprinting' see Berelson, Steiner, op.cit., pp.41 and 43.28 Alan Coddington, in an unpublished paper, A Studyof PoliciesAdvocatedn ConflictSituationsby BritishNewspapers,tudies ten national dailies over a period of two weeks (12 days) to find outhow the kind of solution they recommend for conflicts in their editorials relates to whether theconflict is domestic or international. He finds quite opposite patterns for domestic and inter-national conflicts: both are relatively low where recommendations in terms of 'external settle-ment' (mediation and arbitration) are concerned, but whereas domestic issues rank high on'mutual adjustment' (compromise and reconcili

structure foreign news 1965 - galtung

Documents