In other words, concepts abstracted from actual existence are ideological. But actual existence can never be apprehended without abstraction because actual existence is singular. Therefore all concepts have an ideological drive, as it were, and any concept can be ideological if it is not in some way referred to existence.
There is not one but two term complements to "ideology"--on the one hand "science," and on the other revolutionary activity, or praxis. Marx does not consider that thinking can be anything other than representation: the method of science is abstraction, but warranted abstraction that is not ideological because it remains referred to existence, and in doing so recognizes its own inherent incompletion.
Marx leaves the middle ground between existence and representation vacant, and this is why he has often been misinterpreted as some sort of economic or even technological determinist. This is a misinterpretation because it assumes that there is some sort of pure activity or process that is distinct from thinking, but that is wrong--in fact there is no pure process, it's just that there is always a remainder in any representation. Since representation is conceptual abstraction, at the very least it cannot account for its own singularity. Marx does not want to give a complete representation of a pure process, rather he recognizes the constitutive incompletion of representation as such and thus wants to limit thinking, which is his biggest differend in relation to Hegel, of course. Theory is always referred back to practice, not as something wholly other (unlike for Althusser, who thus gets into an epistemological quandary) but as otherwise incomplete.
What Marx does not consider is the possibility that there may be a non-representational thinking that is precisely concerned with singularity. Marx's thought points to the necessity of (without fully cashing out) an 'existentialism' that does not simply invert the traditional order of priority between essence and existence but (to use an apt but lamentably loaded term) deconstructs the polarity between the two. This would not simply be a method of abstraction, but a hermeneutic approach to the world-forming potency of that which exists.
Ultimately, however, in order to do this a constructive approach, which goes wholly beyond Marx, would also be needed--to think without representing. This is what Heidegger often describes and also, less often, attempts to enact. His most notable attempts at enacting such a thinking are probably his essays that deal with the fourfold.
This non-representational thinking would open up possibilities for thought and action that are non-ideological, in other words it would not be a conceptual precis of what lies in existence but a signpost and guide for a kind of thinking and action (the polarity of which would thus also become entirely questionable) that shelter singularity rather than seizing it conceptually. In other words, this sort of thinking would be neither science nor ideology, nor even distinct from action in any essential manner.
Hence: Marx and Heidegger, Heidegger and Marx.
