sounds like some prime archibabble to me...just improvise a flexible ever-changing definition like Eisenman did and you will be just fine.  fake it till you make it.  

From Of Other Spaces, Heterotopias

(...)

Bachelard's monumental work and the descriptions of phenomenologists have taught us that we do not live in a homogeneous and empty space, but on the contrary in a space thoroughly imbued with quantities and perhaps thoroughly fantasmatic as well. The space of our primary perception, the space of our dreams and that of our passions hold within themselves qualities that seem intrinsic: there is a light, ethereal, transparent space, or again a dark, rough, encumbered space; a space from above, of summits, or on the contrary a space from below of mud; or again a space that can be flowing like sparkling water, or space that is fixed, congealed, like stone or crystal. Yet these analyses, while fundamental for reflection in our time, primarily concern internal space. I should like to speak now of external space.

The space in which we live, which draws us out of ourselves, in which the erosion of our lives. our time and our history occurs, the space that claws and gnaws at us, is also, in itself, a heterogeneous space. In other words, we do not live in a kind of void, inside of which we could place individuals and things. We do not live inside a void that could be colored with diverse shades of light, we live inside a set of relations that delineates sites which are irreducible to one another and absolutely not superimposable on one another.

..................

Having quoted the above, I find that Foucault's further-on fleshed-out concept of heterotopia is not really  defined in relation to a phenomenologically precipitated imagining of spaces as much as to a a functional-taxonomic imagining; Foucault`s conception seems to be therefore derived from that latter flank and the refence to Bachelard, while pertinent on its own grounds, seems to not to be topically tethered to the cited position of Bacheard.  . Further, again im my estimation,  there is a radical difference between heterogeneity and heterotopicality that exasperates further the impertinence of Bachelard here.

What is hetero in Foucalt is tantamount to a virtual otherness that also contains within it contradictions and perhaps is sieved through his own intellectual fetishism (for instance, in regards to spaces of punishment, isolation and so on). What, on the part of Bachelard, is envisaged as being hetero, by Foucault, however, are the imagined qualities that, by virtue of the housed (and I mean that literally and figuratively) imagination, exceeds the bare factual description of the spaces.

Coming across this studio brief by Eisenman, one can devine that heterogeneity is altogether a formal one, even if its bares an aspect of bilocational (or more harshly, confounding)  virtuality - something Eiseman does a lot, I find. Taking a metaphor and disappearing the metaphorical link so as to justify his architecture. Something that, ironically, the parametricists opposed to Eisenman's ways, in their pretense of an objectivism that smacks of pseudo-determinism, commit equally. From that brief:

Alberti first defined space or spatium and more specifically posited that all space is homogenous, meaning that all objects exist within a consistent and calculable medium.  Alberti’s theory of homogenous space first raised questions regarding the nature of space and the difference between surface and spatial objects.  Spatial homogeneity projected the idea of the organism, that all parts relate to a whole and that all objects in space are bound to one another through universal, mathematic relationships.

Jump to today. The digital, in a subconscious attempt to overcome these Albertian principles, has an advantage in its ability to produce an inconsistent multiple, or non self-same repetition. While Alberti’s notational systems transcribed a single design by a single author, computation has the capacity to produce multiple iterations that the designer must choose from.  In Alberti’s case, “the author of the drawing becomes the author of the building.”[2]  Today the computer is the mediator and the generator of the multiple.  The idea of the inconsistent multiple is one possible critique of Alberti, in that it questions the idea of authorship and challenges former ideas of space that were predicated on homogeneity.

If parametric or algorithmic-based design paradoxically lapses back into homogeneous space, how might an idea of aggregation produce the possibility of heterogeneous space, or more specifically “heterogeneity within an intensive cohesion rather than out of extensive incoherence and contradiction?”[3] The studio will attempt to define the possibility of an aggregated project as a means to critique parametric methods of homogeneous space formation.

Aggregation:

One possible definition is the bringing together of similar or varied elements into a non-uniform body, lacking total integration or complete harmony. In this sense, aggregation can be considered the conflict between the consistent and inconsistent multiple as a means to challenge homogeneity. The idea of aggregation will also help define a methodological approach to the studio.  Students will be asked to rethink prior definitions of space formation and raise questions regarding its internal representation.  The studio will also address the distinction between form-finding, which is arguably what computation typically yields, and form-making, which produces form through an idea about architecture.

indeed!