Kramsch and Whiteside take Bakhtin's (1981) dialogism as something quite compatible with complexity theory and postmodern sociolinguistics. The five features they cite as shared by the three are (1) relativity of self and other; (2) timescales; (3) emergentism; (4) unfinalizability; (5) fractals. (pp. 659-660).
Features (1) and (2) seem to me more important than the others. Regarding relativity of the self and the other they say:
[B]oth the self and the other are intrinsically pluralistic, and possibly in conflict with themselves and with one another. Because the I is not unitary, but multiple, it contains in part the other and vice-versa; it can observe itself both subjectively from the inside and objectively through the eyes of the other. (p. 659)
The plurality in terms of one's identity is amplified by layered simultaneity of different timescales. Kramsch and Whiteside cite Blommaert.
'We have to conceive of discourse as subject to layered simultaneity. It occurs in a real-time, synchronic event, but it is simultaneously encapsulated in several layers of hisotoricity, some of which are within the grasp of the participants while others remain invisible but are nevertheless present' (2005: 130) (p. 659)
The plurality in identity and time is indeed best embodied in Dostoevsky's novels. When I had an opportunity (in fact two, in a workshop and a plenary speech) in AILA 2008 to listen to Kramsch, my reading experience of The Brothers Karamazov on my flight to Germany was just recalled by her speeches, both in terms of content and style. After the workshop, I had a very lucky moment to talk to her, and when I mentioned the name of Dostoevsky, she gave me a smile of agreement. You may also love Dostoevsky perhaps because his multifarious style expressively reminds you of the plurality and complexity of the communication in your real life.
In the last section of the paper, Kramsch and Whiteside propose the concept of symbolic competence.
An ecological analysis of these data reveals a much greater degree of symbolic action than is usually accounted for in applied linguistics. Social actors in multilingual settings seem to activate more than a communicative competence that would enable them to communicate accurately, effectively, and appropriately with one another. They seem to display a particularly acute ability to play with various linguistic codes and with the various spatial and temporal resonances of these codes. We call this competence 'symbolic competence'.
Symbolic competence is the ability not only to approximate or appropriate for oneself someone else's language, but to shape the very context in which the language is learned and used. (p. 664)
Kramsch and Whiteside say symbolic competence operates in four different ways: (1) subjectivity or subject-positioning; (2) Historicity or an understanding of the cultural memories evoked by symbolic systems; (3) performativity or the capacity to perform and create alternative realities; (4) reframing. (pp. 664-667).
As I've read Butler's Excitable Speech recently (in Japanese translation), the third and fourth points were well taken.
This paper was really successful in drawing on "insights from complexity theory and post-modern sociolinguistics to explore how an ecological approach to language data can illuminate aspects of language use in multilingual settings (p. 646)."