Epistemic Settlements: 'guarded objectivity'

A related but alternate account of archaeological reasoning has been advocated by Ian Hodder in his push for an interpretive archaeology. Drawing more heavily upon hermeneutics, or the philosophy and method of interpretation, Hodder has proposed a “guarded objectivity” as a solution to the evaluation of archaeological claims (esp. 1991a, 1991b, 1999). Initially drawing upon the influence of the idealist philosopher and archaeologist Collingwood (Collingwood 1948), who was particularly influential amongst archaeologists in Britain in the 1940’s and 1950’s (Trigger 2003:137-7), Hodder emphasized the mutual relationship between the questions which are asked of the material record and the range of ‘answers’ which may be solicited (1991b:33-4). Foreclosing questions en potentia or lacking creative imagination with which to formulate questions would then necessarily restrict the possible ‘answers’ elicited from an engagement with the past. Avowedly anti-presentist, the potential alterity of past lifeways demands the exercise of empathy in order to formulate questions which the archaeologist may not be conditioned to look for (Collingwood (1948:24-5). Hodder (1991a,b) later updated his hermeneutic approach with more recent insights of Riceour (1981) and Gadamer (1975) (see Johnsen 1992). These hermeneuticists emphasized the limitations of ‘empathetic approaches’ in the attempt to anticipate meaningful questions relevant to past texts (or past societies as applied in archaeology). Rather, following the influence of Heidegger (1962) and his grounding description of ‘being’ (dasein) as always-already in existence and therefore incapable of stepping-outside of our existential embeddedness in abstract reflection, these authors stressed bringing all of our subjectivity to the interpretive act (as, contra earlier phenomenologists such as Husserl, we cannot bracket-out our particular understandings and situatedness in order to afford genuine empathetic insight). In line with these philosophical sources Hodder underscores the integral function of the particular context of the archaeologist (social, political, economic, etc.) as s/he ‘meets’ (in Gadamer’s fusion of horizons) the object of archaeological inquiry - equally inextricable from the past context of existence. In comparison to the preceding discussions of Kuhn’s theory-ladeness and paradigmatic change in scientific endeavor, such a position would not unjustifiably be linked to an explicit strong-contextualism.

A greater variance to both the definition of the domain and source-side method of Binford’s processual approach would be difficult to imagine. Where Binford correlated ‘culture’ to primarily extra-somatic means of adaptation to the environment, and the ‘pursuit of the past’ to be conducted by eliminating as far as possible the subjective bias of the archaeologist, Hodder’s ‘paradigm-shift’ returns to a more historical and traditional normative view of culture as inherently ideational and context-specific (versus the ‘universal’, physicalist conjecture). With the methodological corollary being that the achievement of a ‘view from nowhere’ whereby epistemic independence is maintained represents an elusive and profoundly misguided pursuit.

Yet Hodder contends, in response to charges that the affirmation of past and present contexts as a necessary ingredient to the interpretive approach opens the possibility of ‘hyperrelativism’ (eg. Binford 1982a, 1987, Trigger 1989b), that a ‘guarded objectivity’ is possible following an interpretive approach. Such relativism branding logically follows from the apparent reification of the interpreter’s current context, so that any questioning derive from Following Collingwood’s (1978[1939]:30-42) original prescription for a question-and-answer method, Hodder (1991b) outlines the interpretive procedure for avoiding circularity in arguing for statements of the past. In distinction from a stronger subjectivist position, Hodder makes two qualifications by incorporating Gadamer’s reconfiguration of the interpreter-interpretant relationship along the lines of a subjectivity-objectivity dialectic and Ricoeur’s four fold ‘distanciation’ between authors and texts (viz. material culture as treated by postprocessualists). The critical implications of these emendations suggest that interpretation is not a matter of an interpreting subject encountering a ‘foreign’, ontologically distinct object, and that ‘texts’ (or meaningfully produced material culture) are distanced from their authors and their ‘authorial intentions’ (ibid.:11). These two revisions of Collingwood’s formulation of interpretation appear initially antagonistic: the subject-object dialectic suggests a miscegenation of the archaeologist (as subject) and archaeological material (as object), while Ricoeur’s ‘distanciation’ seems to imply a critical distance between archaeologist and archaeological material. But along lines not dissimilar from Popper’s (1961:31) distinction of the context of discovery from the context of justification, Hodder appears to be arguing that the distinction between the context of production of material culture from the process of interpretation affords a measure of objectivity. The contradiction is specious as Hodder is making two different types of claims: one is epistemological (the process of interpretation) and the other is ontological (the constitution of past material). Intersecting the equation with the archaeologist’s attunement to time, Hodder inverts the troubling dimension of time to the strength of his approach. Time, rather than a diminution of the ability of the archaeologist’s capacity to reconstruct the past, actually affords ‘objectivity’ in his hermeneutic method. Because it is distant from present contexts of interpretation, material culture may contravene or disrupt our expectations. It is the very alterity and remoteness of the past which confronts our ideas, and grants independence between our theories and the ‘data’. As he states in reference to meaningfully organized activity of the past: “This patterned organization, distant from its original meanings, has an independence that can therefore confront our interpretations (1991b:13). So while removing the “blindness of its own ideologies” with its self-assurance of unequivocal epistemology-methodology, Hodder wants to “retain from positivist and processual archaeology a guarded ‘objectivity’ of the material ‘other’ that provides the basis of critique through the reality of difference” (bid:12). Hodder hair-splits the usual conflation of epistemological and ontological security. As with the discussion of Binford (above) and the processualists Chapter 2, the ontological status of a past out-there and distinct is taken for granted (an unscrutinized realism). And its ability to confront our hypotheses was assumed. The matter of vociferous debate was ‘how’ to establish a secure epistemology via reasoning and methodological procedure to ensure such a resolute confrontation of theory and ‘data’. Hodder splits these two dimensions, retaining (at least in his later works) ontological realism with its capacity to confront our interpretations. But rather than placing the onerous on the archaeologist’s procedures for ensuring justifiable inferences or deductions (depending on the processual bent), ontology, the ‘material other’, is sufficient for securing interpretation.

The thrust here, as it was with Shanks and Tilley’s (1989) rebuttal to charges of self-defeating skepticism, is to reconfigure the nature of knowing the past in order to maintain a bulwarked position from which to base legitimate criticism. They have altered the terms of the debate inherited from processualists by insisting that a form of objectivity in evaluation is realizable without it having to be framed within analytically derived philosophy. Such a valuable appropriation will be expanded upon later in relation to pragmatic thought.

Hodder departs from what otherwise may initially sound familiar to a hypothetico-deductive trained ear (with the generic scenario of archaeologist confronting ‘data’ via theoretically informed ideas) with his description of the interpretation side of the equation. Here there is little explicit epistemological wrangling of the sort familiar to processualists as Hodder’s crafts his ‘epistemology’ from Continental sources. This takes form in his hermeneutic spiral (Figure 3). In interpretation, Hodder elaborates Collingwood’s question-and-answer procedure, which upheld a stringent division of archaeologist and objects of the past, with the more contextually attuned method of Gadamer’s subjectivity-objectivity dialectic (compare Hodder 1986, and 1991a,b, 1999).

Figure 3.3: Hodder’s hermeneutic spiral of interpretation

In this ‘hermeneutic spiral’, the context of the archaeologist predisposes certain interests and expectations, and so prefigures questions of the past. Such questions are then ‘asked’ of the archaeological record, which is equally contextualized within the socio-political, economic, environmental and ideological considerations of the past. For the archaeologist situated within her interest-informing context, the possibility of inquiring of the past as a tabula rasa is ruled out. “. . .no interpretation is possible until interpretation has begun” (1991a:34). Once begun, however, the process is never ending (or only arbitrarily suspended), as the dialectic means that no definitive hypothesis testing or protocol is given for determining when a question has been decisively answered. Questions lead to partial ‘answers’. But more importantly the answers provided lead to further questions. These furthering-on questions then prefigure, as incorporated into part of the archaeologist’s dynamic context, additional queries of the past. “Question and answer continue in an endless cycle since every question is itself an answer and every answer frames and creates new questions” (ibid.). Disengaged from the notion of a hermetically sealed scientist formulating explicit questions of a mutually distinct past, Hodder parallels insights by science studies into the messy practice of actual research (eg. Latour 1987, Pickering 1995) in highlighting the give-and-take of his hermeneutic approach. A stronger assertion of the give-and-take involves how this spiraling movement eventuates the fusion of horizons or contexts. In engaging in knowing the past, the “perspective and horizon” of the archaeologists changes just as what is known (what tentative answers are solicited) alters the past as it prefigures what is intellectually pursued (Hodder 1991a:34-5). It is not a matter of simply documenting ‘The past’ in a disinterested manner. Both past and archaeologist are altered and transformed in the present during hermeneutic investigation. This enables Hodder to declare that “The past and the present only exist in relation to each other. . .thus present and past, subject and object mutually determine and alter each other” (ibid:31). What may initially be unseemly in this account of ontological and epistemic entanglement is Hodder’s attempt to “… escape dichotomous thinking” which has been sedimented as ‘realist common sense’ in the modernist tradition of analytic epistemology - a tradition which greatly influences archaeology and the other social and natural sciences .

Properly speaking, Hodder’s approach may not be schematized as a ‘hermeneutic vicious circle’, as this ongoing inter-change of question-and-answer, archaeologist-and-past, forms a spiral oriented to the future Figure 3.3. While temporary abatements of the process may stabilize these inter-relationships in the present, it is more accurately a future-oriented undertaking. Once begun, the momentum of the dialectic propels . This characterization of Hodder’s hermeneutic or interpretive approach suggests that knowledge of the past per se is not the full potential of archaeology. Echoing Shanks and Tilley, “…archaeology is not merely a way of working, it is a way of living” (1992:67).

A salient question, then, is how in Hodder’s hermeneutic spiral is evaluation of claims going to happen? A dialectics of interpreter-interpretant moves us far from the processual desire for ‘independence’ between theory and data, or between sets of data and ‘source-side hypotheses’ (Wyle 1993:25). Drawing upon a non-analytical tradition, Hodder has construed a framework which poses inter-dependence between theory and ‘data’ as the very definition of the ‘interpretive act’. In the spiraling process Figure 3.3 theory and ‘data’ successively inform and alter each other. The question of evaluative criteria, raised by Hodder in the Lampeter commentary with his query as to ‘how’ such evaluation is to be achieved, is most acute due to his radical departure from normative (i.e. analytically derived epistemology) accounts. Indeed, deviating from familiar standards in archaeology how is this alternative account itself to be evaluated? Anticipating the crucial need to justify an unfamiliar description of the practice of archaeology, Hodder contends that it is, in fact, what we (consciously or unconsciously) do. “The above account of the hermeneutic exercise describes what archaeologists of any hue do and all they can do” (1991b:35, and1992, 1999:62). Such an appeal to our honest appraisals of what we do provides good grounds for accepting the interpretive approach, but I will argue in a complimentary manner that a stronger justification for a particular framework is based upon pragmatic criteria. Nevertheless, if the contention is that we operate according to hermeneutic dialectics, then we must evaluate similarly as well.

Hodder’s sketch of ‘interpretive evaluation’ draws upon common sense. Succinctly, he inverts Binford’s assessment of the relationship of epistemic independence versus dependence (the ‘horns’ of Wylie’s dilemma) and splits the difference. Additionally, he makes an important (though partial) movement away from the dominant (in the analytic tradition) criteria of ‘truth’ by ‘introverting’ the goal underlying the considerations of independence/dependence. How does this alternate account of evaluation work?

First, as already described, hermeneutic process of interpretation, in its attempt to get away from dichotomous thought, merges subject and object into a dialectic. Neither, as archaeologists or archaeological material, is independent. They merge and co-constitute each other in the question-and-answer procedure. Yet, critically, they do and must retain some measure of ontological independence. Recalling Shanks and Tilley’s (1989:4) statement that objects resist theoretical appropriation, Hodder claims that “the material remains have an independence that can confront our taken for granteds” (1991a:12). Thus archaeological objects may provide ‘answers’ which confound or contradict our expectations formulated by our prefigured questions. Recalling the discussion of Ricoeur’s emphasis upon ‘distanciation’ of author-text, the material of the past is distanced both from its ‘author’ in the past and from its would-be interpreters. So rather than ‘testing’ pre-formulated hypotheses, Hodder describes the ‘fitting of data’ (1999:33-40). “So I do, in that sense, ‘test’ my theory in that I accommodate my theory to data. . .” (1991b:33). In the Archaeological Process, Hodder presents an extended example to demonstrate how this fitting or accommodation works. So, for example, at the site of Haddenham, the archaeologists were expecting a southern ditch as part of the causewayed enclosure. This would seem to fit expectations concerning the symmetry of enclosures, sot that the presence of a northern ditch seemed to indicate the likelihood of a parallel counterpart (Hodder 1999:34-9). Yet, despite excavating trenches which later clearly showed the presence of a much larger ditch with different fill, the ‘expectation’ of observing on the surface a comparably constructed and sized ditch in relation to the northern ditch occluded the archaeologists from initially recognizing the southern feature. Its initial ‘absence’ represented a material resistance to the theory of the archaeologists. Their expectations did not ‘fit’ the observed features of the Haddenham enclosure. The independence of the ditch resisted prefigured expectations and questions, so that a ‘shift’ in the hermeneutic spiral occurred. Accommodating the new evidence, a shift in the perspectives of the archaeologists transpired, so that new, prefigured questions were then asked of the Haddenham remains. So similar to Popper’s (1962) thesis of ‘falsification’, whereby hypotheses cannot never be proved, but only disproved in light of tests against the evidence, Hodder emphasizes only the negative capacity of objects to be independent – to resist accommodation. Moreover, in his claim that the hermeneutic procedure always necessitates initial interpretation or pre-figured questions and expectations, Hodder further parallels Popper’s (1962:46) attack on inductivism. Finally, complementary to Popper and other non-positivists commitment to the fallibility principle, despite contradictions by the objective nature of the past, there cannot be a final confirmation of an explanation. Final arbitration is indefinitely deferred to the future. For Hodder, this is because “there might still be other ways of perceiving and evaluating the data and judging the plausibility of the account. No final verified certainty is ever reached.” (1991b:35).

The capacity of material to retain ontological independence as a critical component shaping the question-and-answer hermeneutic brings up the second criteria of evaluation for Hodder: coherency. The perpetual accommodation or ‘fitting’ requires the (potentially) endless adjustment of the ‘whole’ of the available data to the particular ‘parts’ under question. There must be part-whole coherence. From the example from Haddenham, the ‘absence ‘ of the expected southern ditch (as a ‘part’) did not cohere with the ‘whole’ of the other evidence indicating that if it is indeed an enclosure, then the expectation was that their ought to be a southern ditch to the enclosure. The initial oversight of the southern ditch presented a contradiction to what a part-whole coherence would predict. Obdurate objectivity, as the differing size and construction of the actual southern ditch, disrupted the initial coherency which had been accumulated by the previous ‘data’. A subsequent shift in expectation, driven by the criteria of coherency, raised alternate questions (i.e. perhaps we ought to be looking for a different type of ditch). Which in turn lead to the eventual ‘discovery’ of the southern ditch and provided a sense of coherency with the remainder of the evidence from Haddenham.

Coherency is of course an internal relationship between the parts and whole. Such a focus of ‘fitting’ evidence together emphasizes the organization of ‘data’ and their compatibility. It does not, as distinct from processual accounts, place correspondence between ‘data’ and hypotheses or models as the arbiter of interpretations. As Hodder (1991a:12) explains, “processual and hermeneutic approaches of course differ in their approach to the validation of hypotheses, emphasizing external and internal criteria of judgment respectively. . .” The supposition of coherency being, whether or not external reality may be demonstrably proven to be configured a particular way, reality is organized systematically and consistently. So that plausible accounts of the past ought to be logically coherent and the mechanisms or structures proposed compatible and consistent.

Now, here is where there develops some dissonance in the various components proffered by Hodder in his overall interpretive platform. Foremost is the tension in Hodder’s desire to retain objectivity in his evaluations, even if such a status is to remain a ‘partial objectivity’. While using his hermeneutic procedure explicitly entails the context-dependence of theories as they inform research questions, there remains the goal “that we should be ‘objective’ as we can and that using ‘independent’ theories helps.” (ibid:35). However, Hodder’s form of interpretive archaeology, underscoring the situatedness of all phenomena, including present archaeologists and past individuals and materials, espouses what amounts to a ‘strong contextualism’ in Kuhn’s accounts of scientific practice. Now, some have argued both within archaeology (eg. Binford 1989:34-9, Kelley and Hanen 1988:69-76, Wylie 2002:125) and in philosophy (Davidson 1974) and philosophy of science (Hacking 1983:65-74, Scheffler 1967:20-2), the dependence of theories on paradigms or historical context is only partial, or even, with Davidson, founded upon illogical premises (i.e. the ’very idea of a conceptual scheme’). Such a critique, as we saw with Binford’s solution to the ‘red herring of paradigm dependence’, may indeed afford an argument for ‘partial objectivity’ (indeed, such an line of reasoning will be discussed below with respect to Alison Wylie). But Hodder, after stating the desire for ‘guarded objectivity’, continues in the quoted paragraph to reassert context-dependence central to his position: “But all those procedures can do is increase the plausibility of our accounts. There can be no absolutely objective data or independent theories since data and theory are linked to ‘frames of reference,’ horizons, questions, or contexts” (1991a:35).

I would argue this is a ‘cake and eat it to’ wish on Hodder’s part. But he is not unaware of this internal inconsistency. Tracing the development of his ideas of the possibility of objectivity in interpretation from his earlier works (eg. Hodder 1984:27-9) to his more detailed treatment in The Archaeological Process (eg. ch.3) reveals the attempt to surmount this difficulty of maintaining a ‘strong contextualism’ inherent in hermeneutic philosophy and ‘guarded objectivity’. I summarize this as Hodder’s movement from ‘anti-processualist’ arguments against the possibility of objectivity obtained by positivist reasoning to his pro-interpretive arguments for an ‘objectivity’ obtained by coherence-cum-correspondence. So for example, Hodder initially emphasized that ‘consensus’ was what actually underlain specious claims to objectivity . “. . .theory and data do not confront each other within an objective science of archaeology. . .assumption is built upon assumption and a consensus is reached” (1984:26). More recently (1999:59), though, is the warning that “To reduce truth and objectivity solely to intersubjective agreement is dangerous and inaccurate.” This propels Hodder to support a firmer basis for evaluation. He proposes a supposed integral relationship between coherence and correspondence and their mutually reinforcing ability to ground a form of objective evaluation.

“The notion of correspondence between theory and data does not imply absolute objectivity and independence, but rather embeds the fit of data and theory within coherence. The data are made to cohere by being linked within theoretical arguments. Similarly, the coherence of the arguments is supported by the fit to data. . . .correspondence with the data is thus an essential part of arguments of coherence (ibid:61).

Synthesizing his earlier ideas of ‘data answering back’ in the hermeneutic spiral and the internal (to the horizon or context) coherence of claims, such an updated position well describes what archaeologists do in practice. In principle, the position would draw upon two influential theories of truth (coherence and correspondence) to bolster claims made within such a framework. Interestingly, the two accounts historically developed as competing alternatives. Taken on its own, a coherency theory is more than simply consistency of propositions (though see Davidson 2001:43-4). This is an earlier and weaker version, asserting that, for instance, as long as archaeological interpretations or claims do not contradict each other then individually each claim is probable. Obviously, especially for a field such as archaeology, the ‘data’ routinely, if not as a matter of course, underdetermine claims so that counter examples may easily be imagined. This is particularly true given post-NAGPRA archaeology Chapter 1 which must account for internally consistent frameworks, such as local indigenous knowledges and values of a heritage site and archaeological assessments of the same locale, which nonetheless may be mutually incompatible. As discussed in Chapter 1, this is best instanced in the conflict of oral history accounts and archaeological data (eg. Anyon 1996, Anyon 1997, Echo-Hawk 2000). However, Hodder’s platform of interpretive archaeology avowedly supports a ‘open archaeology’ allowing for the registering and inclusion of the ‘others’ voices’. For instance:

“Interpretive archaeology can be an active, ‘doing’ archaeology. We need to see postprocessual archaeologists launching into coherent and sustained interpretations of the past, involving themselves in whatever contemporary issues those interpretations raise. . . In the United States the relevant debates may concern. . .the interpretation of American Indian remains” (1991b:16).

So to this weak form of coherentism is flawed as it is incapable of maintaining consistency across alternate ‘sets of propositions’ of the past that is explicitly called for by Hodder’s own platform.

Alternatively, what Hodder may be forwarding is a more specified coherence theory (eg. those of Putnam 1981, Young 2001). Meaning, consistency must be specified with respect to who holds the propositions or statements, background knowledge, established data, and so forth of the past to be true or reliably established. Coherentists differ on this point along a spectrum. Briefly, they generally agree that the ‘specified set’ (or this background knowledge) consists of propositions believed or held to be true. They differ on the questions of who believes the propositions and when. For example, is it the current totality of actual archaeologists (eg. Young 2001)? Or may it be all archaeologists once they have reached a (hypothetical) limit of inquiry (eg. Putnam 1981)? Or does it rest on some transcendental notion of an omniscient observer? Given his ideas of the perpetual question-and-answer interpretive process, one that is only temporarily abated, Hodder likely holds a version similar to that of Putnam’s. This influential thesis of ‘internal realism’, in its attempt to avoid the pitfalls of subjectivity and objectivity, holds that justification of claims is accomplished by appeal to an ‘ideal epistemic context’ consisting of the hypothetical ‘end of inquiry’ and an ‘ideal set of knowledgeable inquirers’ (1990:113-17, Putnam 1988:107-16, 2002:20-22). Of course there are problems with such a ‘hypothetical’ basis for specifying who would agree to all of the propositions or claims (eg. Dummett 1976, Rorty 2000). But the critical point is that even if Hodder is appealing to a notion of coherency of this sort, such a coherency bases itself on inter-subjective agreement amongst (however specified or inclusive) a set of inquirers. As we have seen, Hodder vehemently denounces inter-subjective agreement as being “dangerous and inaccurate” (1999:59).

So whichever form of coherentism Hodder is forwarding, both are susceptible to inconsistency when coupled with his interpretive archaeology as a whole. Either consistency of archaeological claims must be accomplished by the very sort of restriction of participation in global heritage which Hodder has been an important spokesperson against, or the specified group of archaeological inquirers must be greatly enlarged to include stakeholders who are equally engaged in ‘epistemological adjudication’, in which case the result of consensual and inter-subjective evaluation must be accepted. In his pioneering direction of work at Catalhöyük (esp. Hodder 2005, 2000) and in his writings on the issue of multivocality (esp. 2002, 2004), Hodder seems to be in fact moving towards the latter.

Linking either of these versions of coherency to the theory of correspondence still begs the question of how a set of interpretations or claims, however deemed ‘consistent’, may be said to irrefutably refer to the things of the archaeological record. As discussed in Chapter 2, archaeology has predominantly, with few singular exceptions, assumed the veracity of the correspondence theory of truth. As counterpoised to a coherency view, the evaluative criteria for accepting statements as justified or true in correspondence rest externally upon the supposed direct relationship between such statements and the independent existence of phenomena corresponding to them. It is a mirroring strategy. Provided our statements accurately describe what we observe ‘out there’ then there is a basis for accepting them as justified. As the mounting arguments against a correspondence theory will be more fully discussed below as central to an alternate, non-foundation basis for justification, I will only consider Hodder’s linking of the two together as a final component of his most recent estimation of evaluation – what he has termed “external coherence” (1999:61). The linkage is formulated in two ways. The first is tacit. It assumes that coherence somehow ensures correspondence. That is, if the interpretations cohere or are logically compatible, then they must refer or be exact in their fit to the evidence (Hodder 1999:61). This assumes, as stated before, that the ‘reality’ of the archaeological remains, as the material outcome of human and natural processes, are intrinsically coherent themselves. So from the Haddenham example, other causewayed enclosures of the British Neolithic in the Cambridgeshire Fens, led to the expectations that the Haddenham site would be consistent in material and construction to other known examples (ibid:33-4). When the initial ‘invisibility’ of the southern ditch challenged this expectation of coherence, the dissonance between expected coherence of material and the visible remains themselves created a dilemma. As Hodder describes it, this assumption of coherence made ‘invisible’ the actuality of the southern ditch – “we still could not see the southern of the two parallel ditches” (ibid:36). Assuming coherence of the material remains inhibited ‘seeing’ the actual southern ditch with its different construction and material. What is problematic is at what level of ‘coherence’ is the underlying human and natural processes supposedly organized. Initially, an expected ‘precise coherence’ in terms of construction and material was clearly incorrect. Broadened to a more liberal level of ‘coherence’ (so in terms of general layout), then the eventual ‘discovery’ of the discrepant southern ditch could be said to substantiate the assumption of coherency of the archaeological remains. This assumption only holds if a very loose and adjustable notion of ‘cohere’ is employed. In which case it can conceivably support any (or at least a range) of interpretive conclusions. It is far from a secure, foundational basis for claiming that coherence must entail correspondence to ‘the facts’.

Secondly, the ability of the real southern ditch to ‘talk back’ against the expectations of the archaeologists is argued, as part of the question-and-answer hermeneutic spiral, to ‘check’ the pre-understandings and theory with objective fact. It is reality interjecting itself into the interpretive process, ensuring that claims, however coherent, correspond to the archaeological remains. On itself, such an admonition would accurately describe how archaeologists continually accommodate evidence as it emerges over the course of time. And tied to arguments against full Kuhnian contextualism, or the pervasive influence which research paradigms (or ‘disciplinary matrices’) hold over identifying, selecting, and constituting data relevant to informing theories, such a line of reasoning for at least ‘partial independence’ of theories and data would strengthen Hodder’s proposal (though see the discussion on Binford’s ‘solution’ to the dependence-independence problem above). It has already been remarked that Hodder’s simultaneous affirmation of context or perspective dependence and the independence of data to ‘answer back’ presents difficulties. This overall attempt at not dichotomizing subject and object through dialectics, or being ‘data-led’ in the hermeneutic spiral, does not avoid some of the same problems as more processual accounts as it begins from the very realms of dichotomous thought (subject and object, dependence and independence) which it seeks to overcome in ‘synthesis’. Incorporating the terms of the debate set by dichotomous structure, the solution still remains troubled by these intransigent concepts of modernist thought. So attempting to integrate a coherency settlement (subject or archaeologist centered) with a correspondence settlement (external or object centered) brings with it the self-same residual problems of synthesis through dialectics. This second aspect of the linkage of coherence to correspondence operates with the appeal to evidence as being capable of negatively disrupting theory. “So we cannot separate the coherence of the argument from correspondence with the data. The latter is constructed within the former. . .the relation between theory and data is thus dialectical. Theory and data depend on but can contradict each other” (Hodder 1999:40). So ordinarily, when interpretations cohere, it is argued that this is so because the external structure of reality, or the evidence, is coherent. This mirroring of coherence as confirming ‘fit’ makes several assumptions about the coherency of nature, and the level at which such a coherency is selected for. But in disrupting coherency of theory, the data seem to be checking an incorrect set of interpretations. But, as Hodder states, correspondence is constructed or determined by coherence. Judgment of coherency is relative to the informing theory. It depends upon at what level or at what scale coherency is selected for, and what evidence is included in consideration. Coherency is not ambiguously all encompassing and universal. Therefore, correspondence, or, in the negative aspect, the lack of correspondence between theory and data depends upon the judgment of coherency. It turns out that even in negative examples, correspondence and coherency are mutually dependent and circular. We cannot get around the questions of ‘coherent’ for who? Or the evidence is not cohering for who? With what standard and scale of coherence, and so what standard for negative coherence or lack of fit? Hodder acknowledges this integral component of judgment in coherence-correspondence: “However, such numerical indication of correspondence always have to be evaluated against contextual relevance and interpretation in order to determine whether the different examples of fit are relevant to each other” (ibid:61). Unsurprisingly, the strong contextualism of Hodder’s interpretive approach raises its head again. And this throws ‘correspondence with the data’ back into the camp of coherence. And coherence, as already discussed, relies upon inter-subjective agreement or consensus of a particular set of individuals. Which Hodder want to move away from as being ‘dangerous’ and, I would venture, non-dialectical as its inexorably tied to subjectivity.

Hodder’s arguments for coherency of interpretations amongst archaeologist, while as I argue, are on their own incapable of avoiding several objections, move us productively away from the defects of processual reasoning and have opened up the necessary reflexive and critical questioning of how archaeology establishes authorial knowledge. Delving into the angst of epistemology is especially important in the contemporary moment given archaeology’s ever increasing role as a nexus for issues of global heritage and indigenous and local involvement. The combination of the ‘external mandates’ Chapter 1 and internal, epistemological angst cannot be ignored or confronted without some soul-searching on the part of the discipline. Indeed, this push towards a radical or democratic inclusion, recalling the ‘communicative epistemology’ of Shanks and Tilley, provides the impetus for the current project. But there is a sense in which the charges of relativism, especially those coming in Renfrew’s (1989) review of Shanks and Tilley’s collaborative work, and in the Lampeter Archaeology Workshop, have impeded a fuller exploration of how knowledge claims may be made in archaeology without worry of the need to secure such claims in a form of foundational epistemology – either coherency or, more generally, correspondence. Hodder, Shanks and Tilley have, in their discussion of coherency as a principle for evaluating claims and most importantly in their early criticism of correspondence, have moved the discipline much closer to such a framework unencumbered by the impossible modernist dream of epistemology. And in some of their later works (Shanks 2005, 1992b, Tilley 1999, 2004) they have either avoided reference to epistemological issues, or (eg. Shanks 2005:148) have maintained a critical attitude to representation and correspondence. It is my intention to carry a step further what I consider to be extremely productive within the ambit of these archaeologists and their collective challenge to foundational epistemology in archaeology.

Forward to Epistemic Settlements: ‘mitigated objectivism’

Containments

Return to Epistemic Settlements: 'partial and contingent objectivity'