Notes on a passage by de Man
The following passage comes from de Man's review of Of Grammatology, (Critical Writings 214-15); it was suggested as a putative description of deconstruction.
... we find a true reading here, not limiting itself to sketching the features of a thematic narrative that can be organized only by ignoring or masking the discontinuities abounding in Rousseau's discourse but, on the contrary, discerning in these very lacunae the principle of a complication raised to the level of a system. Not a literal reading, then, which would be limited to juxtaposing incompatibilities, nor strictly speaking a symbolic, thematic, or even a dialectical reading, which would attempt to reconcile the different discursive levels among themselves. What is involved here is a truly hermeneutic reading, one that traces the contours of a field of signification by means of the logical points of resistance strewn throughout the text, rather than at their expense....I should like to make a few remarks about this passage. Let me begin with that remarkable initial phrase, "we find a true reading here."
"Reading" is lit-crit jargon, one of the standard devices used in the lit-crit game to alter and reconstitute meaning. Thus written material (and sometimes painting, sculpture, and what not also) is "text". In the ordinary course of things one speaks of poems, essays, novels, play scripts, expository text, et cetera. Writers write books, articles, et cetera. What almost all of this written material has in common is sequences of words which ordinary readers would indeed call text. Of course books can and do have material which is not "text" in the sense of sequences of words, e.g., pictures, indices, tables, et cetera.
What is achieved by using this specialized usage of "text"? One thing that is achieved is the erosion of surrounding context. Thus the physicality of the book is erased and classifications such as "poem" are eliminated. What is left is just "text", sequences of words. To this text one can then apply the machinery of literary criticism as though all text were the same. One may well ask, why not reduce written material to character sequences or bit strings? One can of course, but then one is dealing with databases, programming, and such trash. The style of reduction that one uses depends upon the machinery that one wishes to use. I presume it will not have escaped the reader's attention that academia deals with text books.
There is a refinement in this maneuver that should not be overlooked. Commonly one speaks of the text in a book or other containing object. Thus one might speak of the text in this article. The maneuver is to erase the container; thus this article becomes a text.
Texts are like animals; not all texts are equal. In particular, as we see in the passage from de Man, we have "reading" and "discourse". A discourse is or may be the written equivalent of a lecture. In practice it seems to serve as a label for starting material for the interpretive cycle, which brings us to reading.
In less exalted arenas people hold readings, which is to say somebody reads some material aloud and other people listen. Evidently this is not what de Man means by a reading. What is going on here is a theory of thought is being smuggled in via the vocabulary. At the initial level one understands what one reads unconsciously. That is, one reads and understands without a conscious process of interpretation involving the production of new sentences and words that "explain" what one is reading. This is not to say, of course, that one cannot produce such auxilliary interpretation; one can. The nub is that the act of reading (as it is ordinarily understood) requires this unconscious assimilation and interpretation of the material being read.
When de Man speaks  of reading, however, he is not speaking of the physical act of reading; rather he is speaking of a bundle of text that is produced and is represented to be interpretation of and commentary on another text. The physical act becomes invisible; it is replaced by another species of text. Not all bundles of interpretation and commentary count as readings. Thus, "The Lord of the Rings sucks" is not a reading. To be a reading the text purporting to be a reading must in some sense mirror the original text. Generally one has an original discourse and a cloud of readings around the discourse, where most of the readings are commentaries on other readings. 
One of the tricks of the trade in deconstruction game  is to note the presence of binary hierarchies. Extra points are given if one of the alternatives is suppressed. Here we have an example; reading and writing are a pair where "writing" is suppressed, being hidden in the back ground as "text" whereas reading is promoted, but must acquire the attributes of writing. This is a reversal of another hierarchy, original/derived, in which original work is valorized over derived work.
According to the quoted passage  _of Grammatology_ is not only a reading, it is a true reading.  What might de Man have meant by this cryptic "true"? Perhaps he meant nothing in particular - he was simply making approving noises. That may be the safest explanation. It does seem, however, that there is something being said here. If we take "true" as "faithful" we may ask: Faithful to what? The claim could be that it faithful to the original material by Rousseau, or it could be that it is faithful to the how readings should be constructed, or more boldly, it could be a claim that one can only do a true reading if one does it in the true way. This latter claim is supported by the body of the quote as we shall see below.
If we consider the relevant section in _of Grammatology_ to be an example of deconstruction (and this is subject to debate) then de Man's remarks can be read as a claim that deconstruction is the true way conduct the manufacture of readings. Said remarks have been presented as a description of deconstruction. This is true only in the vaguests of senses; de Man's passage has the structure of a sales pitch or endorsement, i.e., "Gooberol is not blah, blah, blah (BOO!), nor is it blah, blah, blah (BOO!); rather it is blah, blah, blah (YAY!)" This is quite reasonable, given that it is a review of a book.
Perhaps the most striking thing about the quoted passage is not that it is opaque but rather that it is badly written. One can speculate that such passages are necessarily badly written. Be that as it may the body of the passage is not particularly obscure, at least until we get to the YAY material.
The first "no no" that is avoided in the reading is being limited to a thematic narrative.  The objection to a thematic narrative is that it can only be organized by "ignoring or masking the discontinuities". Likewise a literal reading, a symbolic reading, a thematic reading, or a dialectical reading, none of these will not do because of the intrinsic incompatibilities in the original. In short none of the usual techniques of interpretation suffice.
We now have arrived at the centerpiece of the puffery. De Man tells us that oG provides a true interpretation. The magic "traces the contours of a field of signification". The term, "a field of signification", appears to be jargon which may or may not mean anything. Considered on its face, it is a dubious metaphor. Fields are multi-dimensional, but they have local continuity across dimensions. Text is one-dimensional and threaded.
The passage explains the magic with "traces ... by means of the logical points of resistance strewn throughout the text, rather than at their expense". This makes sense. The notion is that one looks to the contradictions that are reflected in the "logical points of resistance" in "the text". One can examine what is said by noticing what is not said or even what must not be said.
One can advance the proposition that this tracing of the points of resistance cannot be done mechanically, that one has to follow the text being "read" as it naturally flows. DeMan's comments can be read as a claim that Derrida has "done it right."
Is this passage an accurate description of "of Grammatology"? In my opinion it is not - Derrida is not so much "reading" Rousseau as he is "using" him. The resistances in Rousseau that he focuses on exist but those exploited in oG were chosen by Derrida for his own purposes. oG is a riff on Rousseau rather than a deconstruction of Rousseau. This is my opinion; others may differ.
Does the de Man passage give a description of deconstruction? Not particularly.
 More precisely the section in of Grammatology which rambles on about Rousseau. One supposes that "a reading" is necessarily "a reading of" something, although I can envisage that the something might be vestigal and be present only as a trace.
This page was last updated March 1, 2002.