Chapter 2. Poetry in a political preface
2.3.2.2 Second-order networks
These two parties may be seen as ‘second-order’ networks (after Milroy 1987:
46–47). These are looser or more “heterogeneous” than first-order teams, in that
they involve different people with different interests, and not everyone knows or
interacts directly with everyone else (Venuti 2000: 477). They are also usually larg-
er, and thus have more communication links between actors and greater stability.
This implies that discourses and practices (habitual actions) involving network
members often follow relatively stable, frequently reinforced patterns.
Thus many of the shared discourses and practices which bring about social
conditioning operate at second-order level. As second-order networks link mem-
bers of different teams, however, they may also be activated via first-order teams.
How I translated the Dizdar extract (keeping the target-poem’s semantics very
close to the source text, for example), therefore, will have been influenced by sec-
ond-order conversations within UK poetry-translating communities, critics’ re-
views, etc., but also by first-order input in translation classes when I was a student,
for instance.2.3.2.3 Third-order networks
Some meshes of relations, like ‘Bosnians’ or ‘English readers’, may be called ‘third-
order’ networks. These are so heterogeneous and loose that their grounds for
membership and boundaries are best seen as “imagined” by their members
(B. Anderson 1999) – that is, determined largely by belief and self-image. Thus
some of the essay team members call themselves Bosnian. Since independence in
1992, this means not only that they were born or brought up in Bosnian territory,
but also that they feel allegiance to a Bosnian state.2.3.3 Sketching out the framework
A model of ‘real-audience’ poetry translating, therefore, would put the translator
as cognitive and emotive subject at the centre of a web of relation, communication
and action with other subjects, human-made artefacts and physical objects – a
web which lies in a ‘situational’ context of time and space (Chesterman 2002).
Such a model must also recognise that poetry translating’s action is itself commu-
nication (translating), and that it involves a special type of artefact (poems). A
graphic sketch of the key elements and inter-relationships within a preliminary
model that fits these criteria is shown in Figure 4. Following Abdallah (2005), hu-
man actors are marked by ovals, non-human actors by rectangles, and actions and
relations by arrows.