When working with events from long ago, I am far from always being able to reconstruct my own remarks verbatim. Memory preserves individual words, intonations, the sequence of actions, and the overall meaning of a conversation, but a substantial part of the specific wording is lost over time.
In such a situation, there is a temptation to fill in plausible lines based on how the conversation might have sounded. Yet plausibility does not make a reconstruction reliable. A phrase that is convincing as literature may never have been uttered, and therefore cannot be presented as a memory.
But the loss of exact words does not mean that nothing can be reconstructed.
I can reconstruct not my words, but the conception of myself from which I was speaking at the time.
By a conception of myself, I mean here not an abstract self-assessment, but the subject’s concrete position within what was happening. Who did I consider myself to be at that moment? How did I understand the situation? What did I regard as already proven? What was I counting on? What did I want to achieve? Which actions seemed permissible, logical, or necessary to me? What meaning did I attribute to the actions of other people?
This position can be reconstructed from a collection of traces: surviving messages, subsequent actions, emotional reactions, decisions, memories of how the conversation unfolded, and the broader context of that period. Each individual trace may be incomplete, but together they make it possible to see the internal logic from which words and actions arose.
This method does not reconstruct a lost remark. It reconstructs the conditions under which that remark arose.
There is a fundamental difference between these two operations. In the first case, the unknown is filled in with an assumption. In the second, the unknown is preserved, while the structure accessible to observation is reconstructed around it. I do not claim to know the exact words I spoke. I show the understanding of myself and the situation from which I was speaking at the time.
This avoids two opposing distortions.
The first is literary substitution, in which an absent memory is replaced by a vivid, convincing scene.
The second is the abandonment of inquiry, in which the inability to reconstruct exact words is mistakenly taken to mean that understanding what happened at all is impossible.
The subject’s position lies between verbatim memory and later interpretation. It is neither an exact transcript of the event nor today’s evaluation of the past. It is a reconstruction of the model of reality within which a person acted at a particular moment in time.
Therefore, when describing such episodes, at least three levels must be separated:
- What has been directly preserved in memory or documents.
- What can be reconstructed from actions and context.
- What constitutes a later understanding of what happened.
This separation does not weaken the text; it makes it more precise. It makes it possible to investigate the past without turning reconstruction into observation or presenting today’s understanding as the consciousness of that time.
The primary object of reconstruction becomes not a phrase, but the subject who could have uttered it.
