Marc Bloch states in The Historian’s Craft that the primary characteristic of historical observation lies in the reconstructing of knowledge from tracks or traces:
Whether it is the bones immured in the Syrian fortifications, a word whose form or use reveals a custom, a narrative written by the witness of some scene, ancient or modern, what do we really mean by document, if it is not a “track”, as it were – the mark, perceptible to the senses, which some phenomenon, in itself inaccessible, has left behind? (1953,54-55).
Thus, the trace stands for or takes the place of the past. An indirectness, which according to Ricoeur gives an enigmatic character to history from a philosophical point of view. Something that does not concern historians much, as their priority lies in their engagment with verifying the value of a documentary trace to a past event (TNIII, 143; MHF, 278).
Yet, history based on documents is meaningful, according to Ricoeur, so long as historians, in contrast to collectors of mass data banks, hold on to the idea of a debt “to people of flesh and blood to whom something really happened in the past” (TNIII, 118). He sees the historian as providing a useful and nuanced corrective to collective memory, the final result of mass data collection. In view of today’s huge collections of big data and data mining, it is salutary to remember Ricoeur’s reflections from the mid-1980s.
The concept of traces was something Ricoeur reflected a great deal over in both Time and Narrative as well as Memory, History, Forgetting with the help of the writings of, among others, Bloch, Le Goff, Foucault, Derrida, Heidegger, Bergson and Lévinas. In his notion of historical time, as a connecting bridge between phenomenological time and ordinary time, he sees time as being refigured by historians through reflective instruments, such as calendars, generations, and archives, documents and traces.*
Of these, a trace is accepted as one, only when historians can recognize it as such, by rethinking, re-enacting, and retracing its significance, thus enabling a return to the past (TNIII, 146, 183, 77).
Traces are fragile. They may be altered, effaced, destroyed, made inaccessible or forgotten (MHF, 415-416, 472, 284). They indicate a here and now – of past events or activities of human beings. They mark the passage of time; give a temporal distance. They are datable, and thus inscribed in time (TNIII, 229).
Ricoeur identifies 3 types of traces: the physical, material trace, which along with archived documents and the questions of historians constitute the basis of historical knowledge (MHF,177); the affective, the existential that we experience as impressions through encountering traces; and the cerebral, cognitive traces of memory (MHF, 415, 427).
Traces disturb, by their very presence. With Lévinas**, Ricoeur believes that traces imply an Other, a historical Other (TNIII, 124-126), thus attesting to Bloch’s “evidence of witnesses in spite of themselves” (Bloch, 61).
**Levinas, “La Trace”, in Humanisme de l’autre homme, 1972, 57-63.