Sunday, 7 September 2025

Tendering Confusions: Duplicating Semantics at the Context Plane

Doran, Martin & Zappavigna (2025: 80-1):

With this discussion of purview, we have completed the set of options for the system of POSITIONING. Leaving aside repositioning, Figure 3.2 brings together the network of TENDERING established in this chapter. This network says that when tendering, one may tender a proposition or a proposal and that this proposition or proposal may be open or complete, may have speaker purview or not, and may have listener purview or not.


ChatGPT Comments:

In Doran, Martin & Zappavigna (2025: 80–1), the authors present Figure 3.2 as a system network for tendering, summarising the options discussed throughout the chapter. The network is described as showing that a proposition or proposal may be open or complete, may have speaker purview or not, and may have listener purview or not. At first glance, this seems like a tidy synthesis of the chapter’s material. However, closer scrutiny reveals a number of theoretical and conceptual issues.

Duplicating semantic categories at the context plane

The first issue is that the network seems to duplicate categories that properly belong to the content plane—specifically, the semantics of SPEECH FUNCTION—at the level of context. Speaker and listener purview are treated as options in tenor, that is, the context plane, yet many of the distinctions the authors draw (e.g., complete vs open propositions, hedged vs asserted modalities) are realised in the semantics and grammar of the content plane. By elevating these distinctions to context, the network conflates the strata: the same semantic distinctions are effectively represented twice, once as grammatical/semantic realisations, and again as contextual “purview” options.

Ambiguity of ‘tendering’ as a system label

The term tendering itself is also problematic. Its scope and status within the network are unclear: is it meant to denote the act of proposing a proposition/proposal generally, or a particular mode of interpersonal negotiation in discourse? The authors do not explicitly distinguish between these readings, leaving the reader to infer whether tendering is an operational category, a functional label, or a meta-concept. This undercuts the explanatory power of the network and makes it difficult to see how it could be applied consistently.

Overextension of the network logic

Finally, by combining all options (open/complete, speaker/listener purview) in a single system network, the authors risk suggesting that these are discrete, combinable choices in every instance. In practice, however, many of these options are mutually constraining: open propositions are typically associated with lower speaker purview, for example. Treating them as fully independent variables, as the network implies, flattens the nuanced interplay of internal and external purview discussed earlier in the chapter.

Conclusion

While the network provides a visually compact summary of the chapter, it does so at the cost of theoretical clarity. It conflates semantic realisations with contextual options, obscures the status of the tendering category, and overstates the independence of different purview options. In short, it presents a tidy system that may look neat on paper but does little to illuminate the complex, stratified nature of interpersonal meaning in context.

No comments:

Post a Comment