There are very few differences, as the reader can observe, between the various versions of Mr. Nicolai's basic testimony. However, these versions are far from identical. The differences have to do with the choice of words. (In terms of linguistic analysis, these differences appear with the use of a neutral vocabulary, an evocative vocabulary or a "significative" vocabulary). Naturally, this choice of words may be due to the investigators themselves rather than to the witness.
Because of Mr. Nicolai's imperfect command of French we tend to believe that the differences between the various versions are due to the investigators. For instance, in the version given by an investigator from a civilian group, the text is more literary and more dense: he "is surprised . . . strange object . . . impressed by this unusual sight . . . worried and amazed" and it often refers to preexisting imagery in the mind of the investigator: "he discovers a sort of ovoid object . . . the witness does not see any antenna, no porthole, no opening . . . Iightning speed." This version fails to note that it is the slight sound (whistling) that attracted the witness' attention. In contrast, the narration mentions a displacement of dust when the device leaves the ground, a detail which did not appear in the GEPAN interview. Finally, the drawing given by the witness to that private group is fairly different from that supplied to GEPAN: it is more in line with the "classic" UFO and it is drawn with a surer hand.
In summary, the differences we have noted do not lead to a negative assessment of the witness. His own subjectivity does not seem to have impacted his testimony either on the affective scale (expectations) or on the cognitive scale (existing hypotheses). However, the verbal expression difficulties Mr. Nicolai experiences may have encouraged the investigators to inject their own subjective interpretations into the testimony. The analysis becomes too complex at this point to lead to a precise, detailed conclusion about this single-witness account. We can only state that it is generally consistent.