May 8, 2022

Survey of Ankersmit on Reference vs Representation

Here is another excerpt from the larger project which might eventually have earned me a Ph.D, until I quit the program in March. What follows is approximately 3,000 words surveying the first seven chapters of Ankersmit's 2012 opus. One of many reasons I dropped out is that I cannot stand keeping this to myself any longer. It won't make a huge impact immediately, but I strongly believe it needs to get out of my laptop and onto the internet. Someday, someone important will value this, and explain to the guild of New Testament scholars why it should matter to them. Until then, I hope you all can enjoy it.

 

2.4     Ankersmit 1 of 3: Representation Governs Interpretation

According to narrative theorist and philosopher of history Frank R. Ankersmit, theoretical approaches to meaning, truth, and reference (“the three pronged issue”) have fixated on the philosophy of language, thereby overlooking the aesthetic nature of historical representation [i.e., narrative].[1] After reviewing prominent approaches to “the relationship between language and the world,” Ankersmit concludes that we need to reframe these issues through the philosophy of history.[2] “At issue is how to account for a complex reality in terms of a complex text—the paradigmatic achievement of the historian.”[3]

   To that end, Ankersmit next considers temporality, the narrative construction of “unity and continuity” which defines the scope of the task in historical writing,[4] and demonstrates that change over time is bound to cause “asymmetry” between events and their description.[5] Although unity and continuity is sometimes easier to construct (e.g. biographical storytelling), conceptual periods like the French Revolution or the Renaissance exist only through linguistic reconfigurations of reality.[6] Historical knowledge requires historical language. Paradoxically, however, the historian’s choice of words is not dictated by choice of subject or topic because historical texts do not referentially correspond to their own “exact counterpart in the past itself.”[7] Therefore, the question of whether or not a historical narrative represents past events successfully must be judged by something other than standards of linguistic accuracy.[8]

   His challenge thus defined, Ankersmit begins his argument by distinguishing sharply between “historical interpretation” (hermeneutics) and “historical representation” (aesthetics), arguing that one takes logical priority over the other.[9] Art must be examined to be understood. Opinions must be of something. For example, the visual images in portraits and dreamscapes must be seen clearly before one considers their potential for symbolism.[10] Likewise, historians convey their points of view only while projecting a concrete vision.[11] “Representation precedes interpretation.”[12] Even those who twist facts to justify preconceived biases affirm this logical priority: The “aesthetic truth” of the “object” to be interpreted constrains the possibilities for meaning–making, at least initially.[13] Meanings must be grounded in something.

   Consider Harper Lee’s To Kill a Mockingbird. At the climax of the novel, the child protagonist narrates a violent experience while her vision is obstructed. Later, she overhears her father with the sheriff discretely piecing things together. Explicitly denied information, Lee’s audience shares the child’s task of reconstructing what happened based on clues in the dialogue, and the “correct” reading enhances the meaning of the book’s title.[14] Certainly, failed readings and alternate readings might occur at one or both levels of meaning, but in all cases sense–making depends on world building.[15] Literary reconstruction takes precedence over literary interpretation.

   Regarding fictional story worlds, Ankersmit says critical interpretations may be rooted in “imaginary facts.”[16] Adapting a phrase well worth borrowing, he says, “We read the novel as if it were true, and the failure to do so will make nonsense of the literary text.”[17]

Not being sufficiently aware of what is shared by history and the novel, literary theorists tended to rely too heavily on interpretation; they then forgot about the representational anchors of the text and started to do things with texts that said more about the interpreters themselves and literary theory than about the text.[18]

   Although Ankersmit’s emphasis on the text unreasonably neglects the responsibility of the reader, we can take his point about “representational anchors” as an inadvertent acknowledgement that narrative inference is essential, as I discussed earlier.[19] In the same way, Ankersmit’s “aesthetic truth” is undeniably subjective and his narrative “whole” can only exist if an audience constructs it as such. Granting this perspective, Ankersmit’s “representation” equates to Chatman’s “substance of content,” the important task of reader inference which constitutes a “low level” of interpretation.[20] In turn, this two–level hermeneutic can accommodate Frei’s insistence that the basic meaning of “history–like narrative” is “the events themselves,” while also satisfying the demands of Frei’s critics to support higher levels of meaning.[21] Representations support and accommodate meanings.

   One way or another, interpretations tend to find grounding in story worlds. Indeed, Matthean interpreters ignored the plain meaning of “district” in Matt 2:22 precisely because their thoughts were grounded in unexamined contextual assumptions.[22] Thus, my alternate context (chapter X) and alternate reading (chapter Y) for Matt 2:22 is primarily intended to demonstrate that constructing context is not optional; it occurs subconsciously if not deliberately. It therefore seems prudent to evaluate the story world and its context deliberately before making exegetical decisions, especially with narratives set in the recent historical past of their original audience.[23]

 

2.5     Ankersmit, 2 of 3: Representing is unlike Referring

Having established his central point, Ankersmit proceeds to argue that linguistic theory has failed to differentiate reference from representation.[24] To refer is to “pick out uniquely,” to identify distinguishable objects with descriptive language, using labels and propositions which are subject to verification or falsification.[25] Philosophically, this one–to–one correspondence of words and things is what has typically defined the relationship between truth and reference.[26] However, the semiotic relationship between words or statements and their specific referents is quite unlike the unique relationship between a narrative text and the representation it offers.[27] Certainly, portions of narrative may indeed refer, just as cartographic markings may correspond to specific locations on otherwise roughly drawn maps.[28] Also, the representational text as a whole may in some sense “refer” to its overall topic with a general “aboutness.”[29] However, because representational wholes are aesthetic constructions and past events are not uniquely identifiable referents, there is no direct correspondence between story and history.[30] Narrative does not refer to the past because art does not identify actual objects. Rather, narrative represents the past in the same way that art conveys an artist’s vision.

   Therefore, narrative representation is categorically unlike linguistic reference. Descriptions within narrative can be accurate or inaccurate but their overall depictions must be evaluated aesthetically.[31] Facing a dozen portraits of Napoleon, we would not judge any one of them to be true or false.[32] Although Mark 1:1 distinctly refers to the “Jesus” his audience already knows (whose factual existence was a shared concept in the cultural repertoire among early Christians), Mark’s Jesus conveys Mark’s authorial vision of Jesus. Prior to audience reception that authorial vision is not a shared concept, and repeated receptions engage the same vision; thus, there is no shared cultural code for Mark’s representation to reference. At most, we might say the only referent to which Mark’s representation can refer is Mark’s representation itself.[33] Thus, the statement Mark’s Jesus is secretive could be called “self-referentially or recursively true” of Mark’s representation.[34] However, any potential relationship between (1) Mark’s text and (2) Mark’s Jesus and (3) the historical Jesus is more complex than semiotic correspondence.[35] Thus, representation defies standard linguistic conceptions of reference and truth.

   Understanding this differentiation explains why previous NT scholars have struggled to engage narrative as representation. Even Frei, who decried the conflation of depiction and reference, could not embrace narrativity without abandoning historical truth.[36] Likewise, Petersen’s attempt to harmonize story worlds with the conceptual plane of “direct reference” revealed his assumption that narrative depictions either do or do not have a counterpart in reality (via “indirect reference”).[37] The pre-enlightenment assumptions that narratives must be referentially true or false left NT narrative critics unable to receive Gospel narratives on a purely aesthetic level as historically contextualized representations (which they obviously are).[38] All this confusion amounts to a glaring category mistake. Narrative constructions do not have exact referents in the actual past. Representing is unlike referring. When Matthew’s Joseph fears the historical Archelaus, that event “refers” only to itself: a narratological occurrence that took place within Matthew’s literary configuration. To uncover possible truths of the actual past would require further investigation, as I explain farther below.

   Regarding non-fiction writing in general, Ankersmit says, “reality does not dictate in what style it should be represented.”[39] It is not past events themselves but authorial configurations of the past which require particular linguistic constructions, and this is why the historian’s choice of words is never dictated by choice of topic.[40] Instead, like artists and storytellers, historians determine on a case by case basis “how language and reality are properly related to each other.”[41] Fortunately, representation “can say certain things about the world without actually saying them,” deftly “intimating without asserting,” and thereby enabling truths of the world to become “self–revealing” through a narrative whole and its aspects.[42] In Ankersmit’s project, this quasi–ineffability of representation is precisely what philosophies of language have failed to account for, and why meaning, truth, and reference require redefinition vis–à–vis representation.[43] (From my own hermeneutic perspective, all of this narrative indeterminacy further indicates the inevitability of narrative inference—on which, see above and below.)

   What is most important for Ankersmit is that representations construct meaning; that is, they do not merely depict but also “require the viewer to take a certain attitude toward the subject.”[44] Metaphors like “Renaissance” and “Enlightenment” present ideas and perceptions overlaid upon reality, as does the famous sculpture depicting Napoleon as a Roman Emperor.[45] Caricatures (“distorted portraits”) can sometimes provide “a more profound insight into somebody’s personality than a photo.”[46] The fictional Boo Radley was a mysterious recluse made over into a profound symbol of innocence.[47] In all these examples, the joint-construction of temporality and meaning enables representations to support receptions that can be firmly grounded in the text and yet freely interpreted at the same time, allowing exegetes to infer and contextualize about the story world without simultaneously rendering historical judgment. However, as I will demonstrate next, proper engagement of representation also facilitates a smooth transition from narratological analysis into the subsequent task of historical inquiry.

 

2.6     Ankersmit 3 of 3: Meaning Determines Truth

In culminating his project, Ankersmit finally explores the potential for representation to “reveal reality” in its own unique ways.[48] When language is used referentially the question of truth can be answered by verification, but when language is used representationally there is no such access, no corresponding “objects” or “properties” to pick out or go see.[49] Representations blend meaning and reality into “a seamless whole” (albeit, as often as possible, by leveraging undisputed facts).[50] The projection of meaning onto reality, like the famous sculpture of Napoleon as Caesar, may reveal “representational truths” that provoke complex debates, requiring further examination and discussion of the aesthetic mimesis itself.[51] This dynamic explains why exegesis alone cannot recover reality from narrative, and why even historians themselves “so often mistake disagreements about meaning for disagreements about facts.”[52]

   Nevertheless, Ankersmit insists ongoing challenge and debate is precisely what enables historians to “achieve historical truth” without “irresponsible and arbitrary speculation.”[53] That is, discussions of configured reality reframe debates and drive further investigations of historical possibilities. Did Hayden White successfully discern Jules Michelet’s interpretation of the French Revolution or did White merely demonstrate “what meaning(s) the reader of a historical text may discern” in Michelet’s writing?[54] When younger audiences bring in new contexts to reverse previously accepted interpretations, are they simply wrong or do appeals to authorial intention rather beg the question?[55] In such debates, “meaning determines truth and not the other way around.”[56] After receiving a representation, one must seek truth by framing questions in terms of representational meaning. In what ways is Napoleon comparable to Caesar? Did Hayden White assess Ranke’s bias fairly? What aspects of life in 1930’s Alabama are captured by Harper Lee’s novel? Did King Herod intend to break up his own kingdom? In ancient history, for the most part, we must also inquire about the narrative construction of temporality, the dynamic tensions of unity and continuity over time (which can only be represented because they cannot be referenced).[57] Did Galilee become independent before or after the Passover massacre? How did Archelaus’s accession affect Judean families? Seeking answers to all of these questions takes us beyond hermeneutics, beyond exegesis of the text, and exclusively into the domain of investigation and hypothesis, ultimately challenges us to become historiographers and stake our own representational claims about the actual past.

   In a concluding argument, Ankersmit expands his central hermeneutic insight (representation governs interpretation) to say that all language usage (even referential usage) within a narrative is subject to the overall representational meaning.[58] The larger whole frames the questions of each detail’s meaning (and possible truth), which makes the complete text (however paradoxically) “more basic” than the individual statement.[59] So, for example, in Matt 2:22, one’s view of the overall situation inevitably governs what βασιλεύει might mean, depending on whether or not we align Matthew’s temporal context with a particular historical time frame (as noted previously).[60] We do not actually build overall comprehension by understanding each word and statement separately. One’s view of the narrative world unavoidably influences one’s view of each particular object within it.

   To summarize Ankersmit’s ideas, representation governs interpretation and the meaning(s) thus derived prompt new historical inquiries. The latter procedure must follow the former. Interpretation requires initial credulity, thereby informing a more comprehensive skepticism. Fortunately, these philosophical tenets have been effectively illustrated by the hermeneutic and historiographical approaches of one outstanding historian of Roman Judea.



[1] Ankersmit, Meaning, ix–x., 50–53, 59, 62–3; cf. 34–47 for Ankersmit’s essential identification of historical representation and narrative representation. For relevant portions of Ankersmit’s earlier writing, see F. R. Ankersmit, Historical Representation: Cultural Memory in the Present (Stanford, Calif: Stanford University Press, 2002)., 11–21, 29–74, 197–217.

[2] Ankersmit, Meaning, 1–28, (Chapter One: “Historicism”), explains his purpose in writing but the scope is broad and the complexity is dense; e.g., 8–10, 27 explains that because Heidegger “failed to tally with what actually happens in historical writing,” (not unlike Gadamer), Richard Rorty’s “dreams of a new agenda for philosophy of language,” cannot be fulfilled unless we “translate the historicist theory of [Ranke and Humboldt] into a more contemporary philosophical idiom.” Cf. 26, 116ff.

[3] Ankersmit, Meaning, ix; see also 46–7, 62, 152–3.

[4] Ankersmit, Meaning, 29–47, (Chapter Two: “Time”). Among others (below), Hayden White, David Carr, and Paul Ricoeur figure prominently.

[5] Ankersmit, Meaning, 41. Ankersmit (26, 41–6) says Louis O. Mink was uniquely ahead of his time on this point but that was largely thanks to Arthur Danto.

[6] Ankersmit, Meaning, 44–45, crediting this point to H.M. Baumgartner, following Danto.

[7] Ankersmit, Meaning, 45, again following Baumgartner.

[8] Ankersmit, Meaning, 46–7. One might suggest exact correspondence is always a conceptual illusion.

[9] Ankersmit, Meaning, 48–63, (Chapter Three: “Interpretation”).

[10] Ankersmit, Meaning, 49–57.

[11] Ankersmit, Meaning, 59–62.

[12] Ankersmit, Meaning, 53, 63, 119n.22.

[13] Ankersmit, Meaning, 52–53 (quoting Hayden White’s critique of Derrida), 62–3.

[14] SPOILERS: Scout’s mentally challenged neighbor stabbed her attacker to death, but Sheriff Tate and Atticus decide that “Bob Ewell fell on his knife.” Prosecuting Boo Radley would have been like killing a mockingbird.

[15] For example, audiences may reject Lee’s view about the nobility of falsifying a police report, but that alternative interpretation still depends on reconstructing the falsification itself.

[16] Ankersmit, Meaning, 119, 140.

[17] Ankersmit, Meaning, 119n.22, 119–21. Ankersmit has hardly coined the phrase “as if it were true,” but his use of it here with respect to fiction has been particularly influential to my thinking.

[18] Ankersmit, Meaning, 120–1.

[19] Section 2.3, above.

[20] Cf. Chatman, Story and Discourse, 31 (quoted above), the “low level of interpretation” being “representation.”

[21] See 2.2, above, especially n.36.

[22] See chapter–sections 1.4 and 2.3, above.

[23] Ankersmit, Meaning, 62–3.

[24] Ankersmit, Meaning, 64–86, (chapter four: “Representation”), 87–101, (chapter five, “Reference”).

[25] Ankersmit, Meaning, 65, 87, 99.

[26] Ankersmit, Meaning, 66–7.

[27] Ankersmit, Meaning 80–81.

[28] Ankersmit, Meaning, 88n.2, 93.

[29] Ankersmit, Meaning, 79–81.

[30] Ankersmit, Meaning, 87–95, 101 (cf. 79–81). Cf. 13–14 on “reliability” and “the dogma of Universal History.”

[31] Ankersmit, Meaningf, 66–7. On description (depicting details) versus narration (representing temporality), see also Mieke Bal and Christine van Boheemen, Narratology: Introduction to the Theory of Narrative, 3rd ed. (Toronto: University of Toronto Press, 2009), 36ff.

[32] Ankersmit, Meaning, 68–9.

[33] Ankersmit, Meaning, 80–2, 92, 95; cf. 146.

[34] Ankersmit, Meaning, 81.

[35] Ankersmit, Meaning, 72–3. Up to this point in the book, Ankersmit has used “representation” to indicate “(1)” the historical text, or “(2)” the story world, or both at once. After introducing his “three–place” paradigm (72–3), Ankersmit begins to also use “the presented” or “the representation’s presented” when he means to indicate the “whole” or its “aspects” (as envisioned apart from the text). The paradigm itself aims to help us “avoid conflating (2) [the story world] and (3) [the real world].” Cf. 74–5, 80, 103, 105, 150, where terms and numbering vary.

[36] Frei, Reading, 40, 44; cf. 99, 103–5. See also 2.2, above, notes 25 & 28.

[37] See notes 56, 57, and 58 in section 2.3, above.

[38] See conclusion of section 2.3, above.

[39] Ankersmit, Meaning, 77.

[40] Ankersmit, Meaning, 76–9, 86; cf. ix.

[41] Ankersmit, Meaning, 78–9.

[42] Ankersmit, Meaning, 109–10; on aspects, see 68–74,84–5, 103–8.

[43] Ankersmit, Meaning, 86, 153.

[44] Ankersmit, Meaning, 74; the quote is attributed to Arthur Danto without specific citation.

[45] Ankersmit, Meaning, 73–5.

[46] Ankersmit, Meaning, 98.

[47] See 2.4, above.

[48] Ankersmit, Meaning, 102–25, (chapter six: “Truth”), 126–56, (chapter seven: “Meaning”).

[49] Ankersmit, Meaning, 102–8.

[50] Ankersmit, Meaning, 116–7; Ankersmit (114–6) says the linguistic turn (pre–Rorty) originally engaged problems of language in science, focused on facts rather than meaning, which is why the humanities (following Rorty) struggled between empiricism and deconstruction, i.e., whether meaning is bound to reality.

[51] Ankersmit, Meaning, 116–8.

[52] Ankersmit, Meaning, 116, 116n.21; cf. 56–58, saying that when explanations configure reality the mental model threatens to overlay (or merge with) one’s view of reality (e.g., Gadamer’s “fusion of horizons”) but this amounts to undisciplined thinking; ontology and epistemology should be as distinct in literature as in science.

[53] Ankersmit, Meaning, 113–4; cf. 118, 120, in which the historian’s “poetic genius” is also conveniently preserved by Ankersmit’s refusal to detail methodological techniques.

[54] Ankersmit, Meaning, 137–8.

[55] Ankersmit, Meaning, 139–41; cf. 130–6. Despite the extreme example of Tarasov–Rodionov’s Chocolate (139–40), Ankersmit resists giving ironic power of the new context to reader inference, insisting “the text is bound to an imaginary reality represented by it.” I disagree. Representation is necessarily constructed via audience reception.

[56] Ankersmit, Meaning, 116–8, 125; cf. 136–7, 143–5.

[57] Ankersmit, Meaning, 34, 38–9, 43–7; Ankersmit’s chapter 7 synopsis does not explicitly incorporate material from his second chapter on temporality but the concept of representation being temporally constructed has preconditioned the argument of chapters 3–7. E.g., Mink’s view of the “merely preliminary” (39–40) is re–evaluated as Baumgartner’s “transcendental condition” (46–7). Cf. 115–8. That Ankersmit overlooks temporality in these later chapters is explained by his focus on recent centuries and European history, in which the “object level” of truth Is less in question because many more facts are “ordinarily taken for granted” (115–6).

[58] Ankersmit, Meaning, 154, 153–6. Ankersmit (156) considers this single point his “Copernican Revolution,” and the remainder of the book (157–256) applies this to various issues concerning the philosophy of history (cf. 218).

[59] Ankersmit, Meaning, 154.

[60] Chapter One, section 2.

Recent Posts
Recent Posts Widget
"If I have ever made any valuable discoveries, it has been owing more to patient observation than to any other reason."

-- Isaac Newton