Paul Jalbert
Critique and Analysis in Media Studies: Media Criticism as Practical Action

Introduction

The concern over what counts as analysis in communication studies in general and mass media studies in particular has generated mixed commentary on the part of several reviewers of work which is informed by ethnomethodological insights and which simultaneously explicates the critical expressions intelligible therein. My analysis of the US network television coverage of the Lebanon War in 1982¹ can serve to demonstrate this concern. The treatment of that work by critics both positive and negative has been peculiar. The negative assessments almost uniformly complained that I had produced a 'polemic' regarding the coverage and not an 'analysis' of it. The positive comments indicated that I had been successful in 'exposing' media partisanship, while also arguing that, in so doing, I had transgressed the boundaries of the analytical project and crossed over into the 'critique' of ideological positions. While it is true that critical assessments along ideological lines could emerge as a by-product of the explication of the meaning-options made available to members² in the reports I analyzed, such critical assessments (e.g. of 'bias') were ascribed to me as my assessments of the coverage rather than to possible members with certain specifiable background commitments. This was a curious situation and it encouraged me to become clearer about the methodological, conceptual and analytical enterprise in which I am engaged. The result is this essay.

What Should Analysis Look Like?

The task of analysis of media reportage is to attempt to explicate, to unveil the social constructions that allow people to make sense of the world: the access that I seek to do this, the medium through which I seek to make this available, is through the study of mass communication media representations of events and states of affairs. I am not trying to uncover which 'meanings' particular members actually discern, but to elucidate such meanings as could intelligibly be achieved. These meanings can logically be argued to inhere in actual texts in virtue of their organization etc.; the issue is what is available to be grasped from them. While we do not know how many possible understandings could be derived from a particular text or communication, we can nonetheless make arguments of the kinds of understandings that can be achieved, and, in many instances, for very specific understandings which can be derived. To make such arguments does not mean that the analyst is 'instructing' people as to which meanings are actually achieved or should be achieved; although much of media 'analysis' does exactly that (see Anderson and Sharrock, 1979).

If, in the logic of the explication, it so happens that certain renditions, certain versions of the world, become elucidated and, in doing so, the results look like, have the appearance of being, a critical formulation, that should not be seen as a departure from the analysis. The 'critical edge' is itself an achieved phenomenon on the part of a possible recipient and the analyst. What ends up looking like a critique a fortiori emanating from the explicative analysis should not be cast aside as undercutting an analytical program. Why should a critical implication, inferential option, etc. undercut the analysis? Why should the analyst be accused of engaging in a polemic because some of those critical factors emerge in the analysis? Aside from the very real possibility that the analyst may himself³ be critical of a text he is analyzing - one of the goals of the analyst could be to be critical - the critical implications attendant upon the analysis is not necessarily the product of a diatribe or a polemic. For example, media analysts obviously will adopt any one of the range of existing ethical positions concerning the abortion controversy. Consider the case of one such analyst who characterizes the logic of categorization embodied in the dispute in the following way: wherever a protagonist in the controversy speaks of abortion as a matter of 'murdering babies', he obliterates the conceptual distinctions between 'babies', 'fetuses', 'embryos', 'gametes' and 'zygotes', along with their very different ethical and even theological implications. Moreover, the available categorization of anti-abortion activists as 'pro-life' can be argued to postulate an implicit contrast to a derogatory construction 'anti-life', i.e. what would it mean to be 'anti-life'? Surely, this would be to beg the question. It may very well be argued that such an account of the logic of the controversy already embodies elements of 'pro-abortion' reasoning. However, there is absolutely nothing which precludes an analyst trading upon his own critical resources for the production of an analysis of meaning-options: after all, these are not idiosyncratic to him but rather elucidate exactly the parameters of a critical interpretation, just as an ethnomethodologist has no choice but to draw upon his practical knowledge as common-sense resources in an elucidation of the logic of any such resources. As Turner argues:

At every step of the way, inevitably, the sociologist will continue to employ his socialized competence, while continuing to make explicit what these resources are and how he employs them. I see no alternative to these procedures, except to pay no explicit attention to one's socialized knowledge while continuing to use it as an indispensable aid. In short, sociological discoveries are ineluctably discoveries from within the society. (1974:205)

This also applies to media analysts in terms of their deployment of 'socialized competence' or common-sense resources, not excluding their critical faculties. The issues here are: to avoid explicit advocacy of a contending position within the analysis itself; to be able to identify succinctly the parameters of socially distributed points-of-view with which one may be in personal disagreement as well as those with which one may be in agreement; and to argue for an analysis without presupposing the validity (or otherwise) of the critical viewpoint being advanced. Often, those best able to elucidate the structure of a particular argument may well be among its most committed protagonists. However, rather than construe this as a defect, it should be appreciated as a potential advantage. Unless all we wish to read, study and analyze are treatments of matters sympathetic from the outset to our own commitments and presuppositions, in which case we will forever shield ourselves from an elucidation of our own taken-for-granted ways of reasoning, arguing and using language in general, the only way we can have analytical access to the cultural phenomena by which the concept of 'criticality' can be measured is to recognize its character as an emergent feature of the actual analysis. Moreover, the analyst engaged in such explications should not be accused of distorting analysis for the sake of a 'critique' or a 'polemic'. On the contrary, insofar as a critical posture can itself facilitate analysis of some contested position, argument or text, why should we impoverish the domain of analysis? Max Weber made a similar claim in different historical circumstances in his essay 'The Meaning of "Ethical Neutrality" in Sociology and Economics':

One of our foremost jurists once explained, in discussing his opposition to the exclusion of socialists from university posts, that he too would not be willing to accept an 'anarchist' as a teacher of law since anarchists deny the validity of law in general and he regarded his argument as conclusive. My own opinion is exactly the opposite. An anarchist can surely be a good legal scholar. And if he is such, then indeed the Archimedean point of his convictions, which is outside the conventions and presuppositions which are so self-evident to us, can equip him to perceive problems in the fundamental postulates of legal theory which escape those who take them for granted. (1949:7; emphasis added)

Weber was not endorsing anarchism as a political philosophy in this comment: he was taking stock of the kind of intellectual insight which can be won by adopting such a (critical) vantage-point as a methodologieal device, whether or not the one who adopts it does so for more than a methodological purpose. I propose that our earlier example pertaining to the use of categories taken from a consideration of the abortion controversy is exactly analogous.

Within the context of ethnomethodological inquiries, a related issue has arisen which requires some comment. Ethnomethodologists are required to analyze practical actions, including (centrally) communicative actions, without any commitment to their adequacy, correctness or otherwise. This policy of 'ethnomethodological indifference' is then sometimes held to preclude the kind of methodological device under discussion whereby adopting a critical standpoint toward some phenomenon can enable us to obtain access to aspects of its organization, logic or structure otherwise not readily available. According to Garfinkel and Sacks:

Ethnomethodological studies of formal structures are directed to the study of such phenomena, seeking to describe members' accounts of formal structures wherever and by whomever they are done, while abstaining from all judgements of their adequacy, value, importance, necessity, practicality, success, or consequentiality. We refer to this procedural policy as 'ethnomethodological indifference'. (1970:345)

In the course of analytical practice, however, such a policy has its own logical limitations: in what sense could it be considered 'indifferent', for example, to conceptualize the practice of 'water divination' as a reasoned procedure as distinct from, say, mysticism or irrationality? Could ethnomethodologists 'indifferently' treat astrological or phrenological judgments as 'achievements' of reason in contrast to 'defects' of reason? Garfinkel and Sacks themselves actually include 'divinational reasoning' alongside legal and psychiatric reasoning as if they were unproblematically equivalent:

Persons doing ethnomethodological studies can 'care' no more or less about professional sociological reasoning than they can 'care' about the practices of legal reasoning, conversational reasoning, divinational reasoning, psychiatric reasoning, and the rest. (1970:346)

I do not deny the importance of striving to bracket off assumptively theoretical treatments of reasoning and activity as far as is possible. My only point here is to draw attention to the fact that some phenomena are constituted by 'judgements of their adequacy, value... success or consequentiality', such as memory, the achievement of gender, making a scientific discovery, or arriving at a determination that someone has committed suicide, all of which phenomena have been investigated ethnomethodologically. After all, a recollection is not an apparent recollection, a scientific discovery is not a failed attempt at one, etc. Given these inherent limitations to the implementation of a strict policy of 'ethnomethodological indifference', and given Weber's argument in favor of adopting a critical standpoint as a methodological procedure for identifying and elucidating fundamental assumptions in a domain of human behavior, I see no reason to limit ourselves as analysts to orientations exclusive of critical ones. The only caveat is not to transgress the boundaries between explication and advocacy under the auspices of analysis.

In dealing with media representations of controversies, political phenomena and similar areas of investigation, analysis is to be understood as anti-intentionalist. It is rather conventionalist. There is neither the evidence nor the argumentation provided to support ascribing to actual reporters intentions to bias, to under-report, etc. The only issue is, using the texts and only the texts, the communicated reports and only the reports, as data, how to provide for whatever viewings such materials can be claimed to make available to viewers. Naturally, to a critical recipient, bias or neutrality can be determined, whereas, to a non-critical or counter-critical recipient, these dimensions of assessment do not become relevant. Whatever dimensions of assessment can be made are intelligible/invokable only if these are logical options in the first instance. That is, some reports cannot logically be determined in any way, for any recipient, to exhibit bias of a particular sort, just as other reports for any recipient cannot possibly (i.e. logically) be found to exhibit neutrality in respect of some position. In my work on the coverage of the Lebanon War, I encountered frequently the objection from reviewers that I had tried to demonstrate bias. This is not the case: what I had tried to demonstrate was that a recipient could have logical grounds for discerning bias within a communicative object. Whether he does, or should, is not my concern. Whether he could is the issue, and the only issue, I address analytically. Moreover, if we are looking at what could be understood, and one such understanding has a critical character, then what does it look like? How is it achieved? Surely, this is as worthy of analysis as any other comprehension option.

The methodological issue at stake here is this: a member may view 'bias' etc. in a report only given his background knowledge/belief(s), moral orientations and like commitments. However, against any such background (which can be described as a possible - even a conventional - background for viewing a report), how could this inform an orientation to a text? Surely, we can analyze the options for a characterization any text embodies, along with a description of various backgrounds (e.g. historical knowledge/belief, political perspective, etc.) for its reception, and arrive at a conclusion such as: against background B, text T can, in virtue of its analyzable intelligibility structures/devices, be heard as, inter alia, biased (for or against position P) or neutral. Other dimensions of assessment available to members include: balanced/unbalanced; accurate/inaccurate; partisan/non-partisan; complete/incomplete; objective/subjective, etc. It takes logical, ethnomethodological and linguistic analysis to show 'the how' of any such achieved understanding that members may arrive at.

This approach proposes that the 'text' (the embodiment of categorization practices and conceptual arrangements through them) be the object of inquiry. While this is not a new approach in principle, there have been some orientations to texts which have created problems. For example, due to a Kantian prejudice, a traditional conception has been to think of texts as passive, inert phenomena, requiring the activity of the mind of a viewer to impose structure, intelligibility or otherwise upon them⁴. Against this, a more recent argument has proposed that the text is itself active (Smith, 1982) in organizing whatever orientations a member can logically find in it. However, we should not take this position to mean that we are substituting a model of an active text and passive mind for the Kantian one of an active mind and a passive text: rather, both are active! In this sense, we have outlined a methodological procedure to elucidate conjointly the structures of textual organization and design with respect to the analysis of available backgrounds for orienting to them. (By 'mind' here, of course, we are not involving any Cartesian category of a res cogitans - we could just as well say 'member', 'viewer', 'hearer', 'reader', 'listener', etc.)

The Matrix of Criticality and its Permutations

There are many interrelated levels at which the concept of 'criticality' can operate in our domain of interest. Consider, as a first approximation, the following distinctions (often conflated or overlooked): the non-theoretical recipient of a text can be critical of (a) the text, (b) the producer of the text, (c) the organization for which the producer of the text works, etc. The theoretical critic can be critical of (a) the text, (b), a given non-theoretical characterization of it, (c) a given background commitment on the part of a non-theoretical recipient of it, (d) the background commitment of a rival theoretical recipient of it, (e) the producer of it, (f) the organization for which the producer works, etc. The analyst, by contrast, who restricts himself to that form of analysis which begins and ends with the text, which locates the text at the center of his analytical attention, is never interested in criticizing producers or recipients, their background commitments or organizational affiliations: he is interested only in portraying as faithfully as possible the intelligibility structures and devices inhering in the text as well as the background commitments which interact with any such structures or devices so as to generate a given possible understanding and assessment of it. Insofar as some such assessments can be demonstrably critical of a text (and, by transitivity, of its producer, his affiliation, etc.) they also need to be grounded in the possibilities made available by the analysis of the text. To provide for the logical option of criticality of a text is not eo ipso to concur in that assessment as one that should he adduced. Should we discover, for example, that a given text makes available for someone of a given background commitment the determination that X is the case whereas X is not the case in fact, must we, as a matter of analytical strategy, state that X is not the case in fact lest we be taken to concur in its facticity? Surely not. (The converse can also be proposed: namely that when a text makes available that X is not the case, whereas in fact it is, the analyst must say so as a part of his analysis - again, surely not.) To claim that this is essential for analysis is to commit the fundamental error of arguing that an analyst must be complicit with one specific account of text and viewer as a condition of analysis. The fallacy involved in this, however, should not be confused with the perfectly reasonable proposal that, given a text T, its analysis and a description of background commitment(s) B, position P can be found as generatable by T in its interaction with B, where B may very well be defensible, conventional knowledge (or indefensible, conventional ignorance) of a particular sort, or where P may well be indefensible (or defensible) from other points of view.

Just because an elucidation of some P given some T and B may make some fellow analysts uncomfortable (politically, ethically), this is insufficient reason (a) for their refusing to grant, on technical grounds, the claim that P was indeed a generated option from the interaction of T and B and (b) for their discrediting and/or rejecting an analyst's work. For example, it can be shown that, given a specific background:

B1: Acceptance of an historical account of the West Bank and Gaza Strip which depicts it as an unjustified occupation of other people's territory.

a specific text:

T: 'The new arrivals from the Soviet Union were taken to their homes in Judea and Samaria to enjoy their new-found freedom from anti-semitism'.

can generate position:

P1: 'T is biased in favor of the Israeli official position'.

Now, consider the gestalt opposite. Holding T constant, but varying B, T can generate a different P. Thus, with background:

B2: The West Bank and Gaza Strip were historically and rightfully Jewish under the biblical names of Judea and Samaria.

text

T: 'The new arrivals from the Soviet Union were taken to their new homes in Judea and Samaria to enjoy their new-found freedom from anti-semitism'.

can generate

P2: 'T is simply a factual account of the situation being described'.

Both positions⁵ can intelligibly be generated from the same text given the difference in background commitments. How is this possible? That is a matter for analysis to reveal. It can proceed to do so by explicating the internal structure of the text. For example, given that any location can be described with more than one set of categories, what procedural basis can he found to animate any given selection? Correctness is not enough, since many alternative depictions can have a claim to correctness (e.g. 'West Bank and Gaza Strip', 'the (Israeli-)Occupied Territories', 'Judea and Samaria', (etc.). Thus, any given selection can be found, varying B, to have been differentially grounded. The analyst does not have to be complicit with one or another of these positions or their associated background commitments to explicate it and to connect it to the broader theme of categorial logic in account production.

From the above argument, the following corollary may be drawn: it is not permissible to ascribe the deliberate production of any P from T by an author (A) or reporter (R). Although commonsense would tell us that there are some reporters who can determine the possible range of background assumptions on the part of their constituency in order to manipulate or manufacture their consent to some particular position in virtue of the way in which they design their account, nothing in the above argument could warrant any such determinate attribution.

'Criticality' is as much a part of commonsense reasoning as 'conformity', exercised or not. Members routinely engage in commonsense critical appraisals of wide varieties of ongoing perceptual phenomena. Their criticality can be achieved at both the tacit and witting levels. One legitimate objective of analysis is to demonstrate how critical viewings are achievable. In order to clarify the distinction between what I shall call partisan common-sense critics, on the one hand, and the explication of 'options for criticality', on the other, I have selected to contrast the treatment by Joshua Muravchik of US network coverage of the Lebanon War with my own analysis of that coverage.

Some Contrasting Conceptions of Media 'Criticism':
A Case Study

In 1983, Muravchik and I, independently and for very different purposes, had occasion to use the resources of the Vanderbilt University Television News Archive in order to study US television network coverage of the Israeli invasion of Lebanon which occurred in 1982. Muravchik published his report in the journal Policy Review (23, Winter 1983), and I presented my (more extensive) analysis to Boston University in partial satisfaction of the requirements for the PhD degree (Spring, 1984c). Among the many points of contrast between our respective treatments of the same corpus of media materials were the following: my analysis was based upon my reproduced transcriptions of substantial segments of the actually appearing coverage; my analytic effort employed techniques drawn from logic, linguistics and ethnomethodology; and my objective was the elucidation of the meaning structures of the media depictions of the war. Muravchik's study, on the other hand, was an exercise in ideological criticism whose aim was to argue that US network coverage could be generally characterized as having been anti-Israeli in tone and substance⁶. His direct quotations from network media coverage were selected to enable him to make this argument⁷. Among his conclusions was the remark that:

None of the networks achieved... (a) rigorous standard in reporting on the events in Lebanon. CBS, however, seemed to be trying the hardest, and it succeeded a good part of the time. ABC's coverage was erratic; NBC's gave the impression that the network was on a crusade. (Muravchik, 1983:41)

I read Muravchik's discussion with great interest: it was the only serious and relatively thorough assessment of the materials upon which I had been working that I could find in print. My own ideological conclusion, however, was diametrically opposite to Muravchik's. This discrepancy posed for me the following intellectual issue: was there a methodological procedure that could be used to decide between our competing conclusions? After all, both of us could hardly expect to represent more than a fraction of the materials we had examined: what criteria of 'representativeness' could be proposed for our respective 'samples'?⁸ Was a logic of 'sampling' relevant to this issue at all? Were our divergent assessments primarily a function of our different political/ethical starting-points, or, as I continue to believe, were they more a function of our differential sensitivities to the logic of the detail of our common corpus of empirical data?

Let us consider the issue of 'sampling' from a technical point of view. A parallel consideration has been addressed in connection with conversation-analytic methodology. How are we to treat the transcribed data extracts which analysts reproduce for explication - are they to be construed as samples or in some other way? Coulter has proposed that:

The options that are actually adopted in any given conversational interaction cannot be assigned probability frequencies for the simple reason that the universe of possible instantiations of any given sequence class is unknown (and unknowable: it is indefinitely large). Consequently, the study of any particular instantiation of a sequence class (e.g. arguments) must be motivated not by an interest in empirical generalization to the class but by an interest in the a priori relations between illocutionary actions. (1990:182-3)⁹

If our interest is not statistical, but 'logical', then our reproduced data sets are not samples but perspicuous instances.

...I do not have to see n games of chess in order to grasp the a priori relationship between the two pawns of the same colour in the same column and the act of having used one of them in an earlier move to take a piece. Given a knowledge of the concepts in question, or of the game in question, and the datum that speakers and players are abiding by the rules of proper use, then one can see that two or more contingent states of affairs are conceptually connected. (Coulter, 1983:371)

Both conceptual and conversation analysis, then, must invoke a model of 'grammatical' rather than statistical inquiry as their epistemological rubrics. Their interests are logical and conceptual, although founded upon instances of actual usage, and not purely empirical in the sense of an inquiry into mere 'contingent regularities'. Media analysis involving the elucidation of meaning structures is also, necessarily, a model of conceptual inquiry in an extended sense. The rules governing category selections, presuppositions and implications, practical inferences, sequential organization, utterance design and other phenomena of technical interest are derived from, and are answerable to, empirical data in the form of actual cases of, among other phenomena, media reportage. However, the adequacy of such elucidations is, with few exceptions, not decidable with reference to the sheer frequency of empirical instances.

For Muravchik, however, there is no methodological scruple involved in his selection of instances, no sense of the problematics of either sampling or conceptual analysis, and no constraint adhered to (or even recognized) in the derivation of critical inferences. Observations such as the following naturally lend themselves to such scrutiny:

Except for the reporting of Alan Pizey and Bob Simon from Beirut, CBS tended to avoid the tendentious or loaded wording that was used often by Peter Jennings, Barrie Dunsmore, and Mike McCourt on ABC, and by almost everybody on NBC. (Muravchik, 1983:43)

How does Muravchik know (or even claim to know) that CBS 'tended to avoid' loaded wording? How often did they 'load words' as contrasted to using 'neutral language'? What could constitute 'neutral language' by reference to which a characterization of 'loaded wording' might be justified? According to what evidence are we to hold the entire network accountable for the instances of 'loaded wording' putatively detected by Muravchik in some of its reporters' coverage? Did these reporters intentionally, deliberately, set out to load their words, as it were, or is there something else at work which could account for their categorial selections?

Muravchik is perfectly entitled to his 'opinions' about US network coverage of the Lebanon War. The question that faces us is to develop methods whereby we can maintain a principled distinction between an 'opinion' (or a set of opinions comprising a 'polemic') and a critical inference grounded in a logical analysis of the same corpus of empirical materials. In order to give this distinction some flesh and blood, I have selected for discussion a fragment of news reportage from NBC television news on Saturday, 12 June 1982. Remember that, for Muravchik, 'NBC's (coverage) gave the impression that the network was on a crusade (against Israel)... CBS tended to avoid the tendentious or loaded wording that was used often... by almost everybody on NBC'. Here is the fragment:¹⁰

NBC News, Saturday, 12 June 1982 (18:35:30)

(Paul Miller in Nabatiyeh)

While Israelis patrolled the streets looking for terrorists, they also looked for ways of winning over the Lebanese people. In small villages like this, the Israelis and their Christian allies are being welcomed. (18:35:40)

It will not suffice to complain against Muravchik that Palestinian guerrilla forces were frequently categorized as 'terrorists', a designation not once employed to categorize the Israeli Defence Forces. What has to be shown is how such a categorial selection was actually used and in what discursive environment. Editorializing commentaries, manifestly direct quotations of Israeli politicians, officials or spokesmen and similar partisan contexts are domains within which such a derogatory characterization of the Palestinian forces as 'terrorists' may be encountered routinely without generating an impression of biased news reportage. (It should, of course, be noted that the sustained and cumulative usage of this category in reference to only one party to the conflict may well have generated prejudicial attitudes on the part of some viewers¹¹. In the extract reproduced above, however, the category is embedded within a live news report by a reporter (not by an editor or network administrator). He is not engaged in direct quotation of an Israeli or anti-Palestinian spokesman. Nonetheless, I claim that his use of this category, according to a logic which I will specify, constitutes grounds for the attribution of bias against the persons so categorized (the Palestinian forces). What is this logic?

Without having to enter into any debate about the correctness or incorrectness of the category selection according to criteria of a moral or political nature, it can be argued that the design of Miller's report affiliates him to the official Israeli (or pro-Israeli) political position¹². Whether or not the category 'terrorist' is, by some standards, a reasonable one to invoke, is not the issue here. The issue is, rather, that the category is a 'disjunctive' one in the technical sense that while it may be ascribable, it is not selfavowable¹³. Of course, this is not to argue that there is a universal rule preferring self-avowable categories to ascribable ones in cases of conflict. If such a rule were in force, we should all be constrained to characterize Joan of Arc as a visionary and not as an hallucinator and the Stalinist Gulag as a People's Democracy and not as a totalitarian abomination. It is clear that many categories from disjunctive sets are specifically not self-avowable by those being categorized. However, a selection from a disjunctive categorial pair (or set) which privileges an extrinsically ascribable category over a self-avowable one (e.g. which selects 'hallucinator' over 'visionary', 'terrorist' over 'freedom fighter', etc.) where the ascribable category is specifically not self-avowable (irony apart) observably commits the categorizer to a contrastively unsympathetic perspective on those being categorized¹⁴.

Consider, further, the availability of an option in this instance for marking the categorization as de dicto - i.e. 'While Israelis patrolled the streets looking for those they call terrorists...' contrasts with the de re characterization actually produced by Paul Miller, i.e. 'While Israelis patrolled the streets looking for terrorists...'. Juxtaposing de dicto with de re categorizations can be consequential for certain contexts and purposes. In this context, neglecting to mark the selection of the non-self-avowable category as de dicto presumptively establishes it as de re¹⁵. Such a presumption links the reporter's perspective on these events to that of one side of the conflict at the expense of the perspective of the other. This holds whether or not the reporter intended to display, was conscious of displaying, or would have admitted to, such an affiliation. As I remarked earlier, my analysis is non-intentionalist in character. Muravchik, on the other hand, employs intentionalist and motive-ascribing formulations at every turn.

Although I do not want to belabor my treatment of this argument, an additional point about it is worthy of mention. Mr Miller spoke of 'the Israelis and their Christian allies...' The category 'Christian' was indeed the self-avowed one for those whom Palestinian and their allied Lebanese combatants referred to with a variety of non-religious categories, among which were 'Phalangist' (the English translation of their political affiliation deriving from their historical relationship to Franco's Fascist party), 'Isolationist', etc. It could be argued that the category 'Christian' was simply the recognitional preference for an American audience, given the embeddedness for the alternatively correct categories within a relatively less perspicuous political frame of reference. The problem which this selection engenders, however, is that it tacitly projects a religious rather than political-military explanation for the minority Lebanese alliance with the Israelis, and one with which many otherwise disinterested Christian Americans could be expected to identify.

Muravchik singles out John Chancellor of NBC News for special opprobrium, referring to his 'crusade' against Israel's invasion of Lebanon (1983:54)¹⁶. However, it was clear that Chancellor accepted consistently fundamental (pro)-Israeli assumptions in his editorial commentaries. Likening the Palestine Liberation Organization to 'Al Capone's mob', he remarked that 'It is inconceivable that Israel would deliberately get itself into a situation like that (the siege of West Beirut). . . What is much more likely is that Israel blundered into it and now is stuck with it' (NBC News, Friday, 6 August 1982, 17:51:00 to 17:52:20). Three days later, Chancellor is explaining to his audience how so many Palestinians came to live in Lebanon. Ignoring the Civil War in Jordan in 1970-1 Between Palestinian and Jordanian forces which resulted in a massive influx of Palestinian refugees from Jordan to Lebanon, Chancellor focuses upon the first wave of Palestinian immigration to Lebanon in 1948. Characteristically for this conflict, as for most others, the events of 1948 have already been conventionally articulated in a disjunctive category-pair, 'War of Independence'/'War of Conquest'. The former is the official Israeli category preference, the latter is that of Palestinians¹⁷.

NBC News, Monday, 9 August 1982 (17:50:40)

(John Chancellor Commentary in Lebanon)

Some of these people (Palestinians) came here to Lebanon in the Israeli War of Independence. That's how the Palestinians became refugees.

(17:50:50)

Chancellor's usage affiliates him to the official Israeli position, hardly the perspective of an anti-Israeli 'crusader'.

D'Arcy (1963: Ch. 1) analyzed a mode of elision in the construction of factual accounts which is pertinent here. For example, I may truthfully say of someone that he entertained his friends at the party, but I will be found morally culpable of misleading my interlocutor if it turns out that the entertainment consisted in torturing someone. I cannot be found to have lied: my 'sin' is rather one of omission. Similarly, were I to describe the Nazi physician Mengele simply as having 'conducted medical experiments', it is not that I have lied but rather that I have elided morally significant information. We do not need to ramify such examples in order to establish the validity of D'Arcy's 'Non-elidability Principle' for morally consequential accounts. Did Paul Miller, and those other commentators who selected the category 'Christian' to describe Israel's allies in Lebanon, violate this principle? Muravchik does not even raise this question. The issues that Muravchik does raise might have otherwise been analytically interesting had he deployed a methodological framework within which to explicate them. I contend that this is unfolding in these methodological discussions and explicative analyses of actual texts in such a methodological apparatus, one which is capable of transforming an otherwise polemical orientation into a critical analytical attitude.

Background, text, position

Among the reasons I have for objecting to Muravchik's style of media critique are his choice of instances of reportage as a matter of his argumentative convenience alone, his failure to motivate his own readings of his materials by anything beyond his own initial preconceptions about the Lebanon War, and his neglect of the problem of ruling out alternative readings of his preferred instances. I cannot claim that anyone in media studies has yet been able to produce a theoretically definitive account of the structures and devices whereby messages are conveyed, impressions created and inferences facilitated. However, earlier in this paper I outlined a framework for beginning to work out some of these issues, and now I shall try to indicate to what extent it may succeed.

In order to proceed cautiously, I shall consider Muravcik's treatment of a brief fragment from the ABC News coverage supplied by its State Department correspondent, Barrie Dunsmore. Muravchik had earlier characterized Dunsmore as someone 'who repeatedly went out of his way to work little digs at Israel in his stories' (1983:49). Whether or not this assertion can be justified is not relevant here: I include it only as data to illustrate a manifest component of the Background Knowledge/Belief Commitments attributable to Muravchik as a viewer/listener. Here is his treatment of Dunsmore's report:

On 14 June, Mr Dunsmore reported:

Lebanese police said today that as many as ten thousand people may have been killed in the fighting. To deal with such casualty figures Israel seems to be gearing up a campaign to justify its actions. Israeli sources told ABC news today they had captured hundreds of tons of weapons and documents and have dealt international terrorism and extreme blow.

There were two pieces of news here. One was the Lebanese police casualty figure which deserved to be reported and to be treated with a degree of scepticism. The other was the Israeli announcement of some of the documents and equipment they had captured. Mr Dunsmore combined the two in a way that negated the second story by implying that it was nothing more than a part of an Israeli effort at self-justification. As the same time he implicitly confirmed the fallacious Lebanese police statistic by saving this was the cause of Israel's need for self-justification. (1983:49)

Using our analytical scheme - Background, Text and Position - to organize our discussion, we can treat the direct quotation of Dunsmore's report as constituting T and Muravchik's commentary upon it as constituting P. The main elements of P are:

1. The Lebanese police casualty figure cannot be trusted - indeed it was fallacious.

2. Israel's announcement of its capture of weapons and documents was independent of the announcement of the casualty figure by the Lebanese police.

3. Given the dubious or fallacious nature of the casualty figure. Israel's announcement could not have been an effort at self-justification.

4. Israel's claim to 'have dealt international terrorism an extreme blow' was correct and not merely a function of some 'need for self-justification'.

The issue for analysis now is to specify how T and P relate within a structure of reasoning. And the question to be posed is: is that structure coherent independently of B?

Consider first Muravchik's inattention to Dunsmore's use of model qualifiers: 'Lebanese police said... people may have been killed in the fighting' and 'Israel seems to be gearing up to a campaign...' Because modal qualifiers can be strategically used to protect, by preserving deniability, an account-producer from the fuller implications of a knowledge-claim, they can sometimes be challenged by an account-recipient, but not disattended. Muravchik must have made up his mind on the basis of other, undisclosed evidence of Dunsmore's anti-Israeli partisanship that he 'implicitly confirmed the fallacious Lebanese police statistic': the logic of the text itself provides no confirmation whatsoever of this attribution to Dunsmore. Indeed one could argue that Dunsmore's use of modal qualifiers itself built into his account Muravchik's desired 'degree of skepticism'.

Muravchik claims that Dunsmore's combination of the Lebanese police report and the Israeli announcement 'negated' the Israeli claim. His basis for this way of hearing the report is simply to assert that because Dunsmore had characterized the Israeli claim as a 'justification', its status as correct or true had thereby been undermined. This is singularly strange, since presumably things which count as justification can without contradiction be correct or true in their factual status. To say something counted as a justification is not thereby to diminish its true. Of course, people can (and often may well) produce accounts in justification of their actions which are false, but this does not mean that justification have that property as a universal feature. Quite the contrary: the more truthful the account, the more effective the justification it accomplishes.

On what basis does Muravchik deny the possibility that the Lebanese police statistic (reported by Dunsmore as tentatively given) was accurate? Why does it 'deserve to be... treated with a degree of skepticism'? For some answer to this, we must consult Muravchik's already manifested Background Knowledge/Belief Commitments. Muravchik is openly pro-Israeli and defended Israel's actions in Lebanon. He has expressed serious misgivings about Dunsmore's attitude to Israel earlier in his article, quoted above. From this vantage-point, we can begin to develop an answer to our question. If we partition the antagonists in the Lebanon War simply as 'Israelis' and 'Arabs', then, since the Lebanese are Arabs, and Arabs are in conflict with Israel, any Arab - including any Lebanese Arab police officer - who produces a war-related account is immediately open to the charge of exaggeration, lying, self-serving bias, etc. However, Muravchik does not give any basis for his inclusion of the Lebanese police force within the set of actual opponents of Israel. No evidence is given to implicate this police force in fighting Israel's armed forces, in expressing anti-Israel sentiments as a matter of official policy, nor in anything other than a neutral posture. (In fact, from other sources one can develop a view of the Lebanese police force which may tend to align them against the PLO and their Lebanese militia allies.) Thus, Muravchik's critique of Dunsmore's text is based, in this respect, exclusively upon incorrigible commitments already held: after all, if one did not hold such views, the Lebanese police report, however tentatively articulated, might have led one seriously to question Israel's use of such deadly force in Lebanon.

I do not want to be understood here as holding that Muravchik had no right to hear/read T in the way he did. From a purely political point of view, I am unsympathetic to virtually every aspect of his Background commitments. From an analytical point of view, the validity of our respective political positions is entirely irrelevant. The issues involved in the critical analysis of media materials are conceptual and methodological: otherwise, the field degenerates into an extension of the very political conflicts whose representations it is trying to understand. In their article 'Biasing the News: Technical Issues in "Media Studies" ', a well-known critique of mainstream media studies, Anderson and Sharrock comment:

We are not trying to argue that one cannot arrive at the kinds of conclusions that media scholars reach. We argue only that these conclusions are not necessarily to be drawn from those (textual) materials, and that those conclusions are not the only ones which can legitimately be drawn from those same materials. (1979:367)

In developing the kind of analysis exemplified here, the constraints which Anderson and Sharrock recommend are accepted. Only logical grammars can have a necessitarian status. The analyst is precluded from stipulating a definitive content to any given hearing/reading/viewing for any given text. The purpose of analysis, including especially critical analysis, is to elucidate the structure of possibilities derivable from the interaction of any given B, T and P and thereby to indicate whatever arbitrary contrivances and logical limitations are to be uncovered within the exhibited reasoning which links B, T and P.

Concluding Comments

I have been arguing for wider recognition on the part of professional analysts of media materials of the existence of criticality as a practical activity and not just as a theorist's privilege. Ethnomethodology has taught us to recognize the constructive presence within the social world of pre-theoretical hermeneutics: it has so far had little to say of the presence of pre-theoretical critical orientations¹⁸. This may well be explained, at least in part, by a general adherence to a restrictive interpretation of the policy of 'ethnomethodological indifference'.

In the preceding discussion, I have tried to show how an interest in the workings of criticality may be investigated in the domain of media studies. This required the formulation of an analytical methodology and a detailed case study involving its application. In order to demonstrate its relative cogency, I sought to specify the results of its application to a corpus of materials and to contrast them to a substantive media critique (that of Muravchik). Further, I tried to show how Muravchik's critique embodied the elements of a proto-methodology which was too weak to sustain his own critical observations. Among these elements were his reliance upon a primitive mode of sampling selection, his preference for stipulating determinate meanings instead of explicating a logic of meaning-options and his penchant for biasing critical inferences almost exclusively on what, from an analytical point of view, are arbitrary preconceptions of a political kind. This is not to say that preconceptions are avoidable, only that they must themselves be treated as phenomena, as part of a domain of data.

Just as no ethnomethodologist, linguist or logician can work with materials in a state of amnesia regarding his common-sense resources and institutions, neither can effective explicatory work be accomplished on political, ethical or other contentious human constructions if the analyst is required to put aside in their entirety his ordinary critical faculties, including his own ideological and/or moral commitments. A recurring problem has been to circumvent the criticism which awaits him at every turn, namely that he has transgressed the boundary between explication and ideological evaluation, or between a critical analysis and a political polemic. I wish to claim that, while there are no guarantees against being systematically misunderstood, the approach being recommended here does further than others of which I am aware in addressing itself to a resolution of this difficulty.

[ Paul L. Jalbert, "Critique and Analysis in Media Studies: Media Criticism as Practical Action", Discourse and Society, 1995 vol. 6(1), pp. 7-26 ]

Webmaster: Slobodan Markovic (twiddle@eunet.yu)

All rights reserved. Belgrade Circle Journal encourages the reproduction of material appearing on its pages, provided that the source and the author are cited, except in cases where this would constitute violation of copyright held by other organisations or individuals.