16.1 Further Process Models: Seleskovitch, Gish, Ingram & Cognitive Load

Chapters 3 and 4 covered Cokely's sociolinguistic model, Colonomos's Integrated Model, dynamic equivalence, and the Demand-Control Schema. Those are the four the exam most often names. But the NIC Role Delineation Study requires knowledge of the "process of interpretation" and "basic research and practice in interpreting" generally, and several further models give you analytic vocabulary that the case studies reward.

1. Seleskovitch and the Interpretive Theory of Translation

Danica Seleskovitch, a conference interpreter and researcher at ESIT in Paris, developed the théorie du sens — the Interpretive Theory of Translation — from the late 1960s. Its central claim, drawn from observing simultaneous conference interpreters, is the deverbalization hypothesis:

1. Comprehension — the interpreter takes in the source and fuses it with prior knowledge.
2. Deverbalization — the linguistic form is discarded; what remains is a non-verbal sense.
3. Reformulation — the sense is re-expressed in the target language's own forms.

The practical consequence is the one every interpreter education programme repeats: you cannot interpret what you have not understood, and you cannot produce natural target language while still holding the source's words. Colonomos's Integrated Model, with its "concentrating / representing / planning / generating" stages, is a descendant of this line of thinking applied to signed-language work.

The common critique is that deverbalization is hard to demonstrate empirically and understates how much form-based processing occurs under time pressure — which is exactly what Cokely's miscue data shows.

2. Gish's Information Processing Model

Sandra Gish developed an approach aimed at a practical problem: interpreters who process at the sentence level lose the speaker's larger structure. Her model directs the interpreter to work top-down, from the discourse's organizing frame downward:

• Goal — what is this speaker trying to accomplish overall?
• Theme — what is the governing topic?
• Objective / Units — what are the major sections?
• Details — the supporting specifics.

Working top-down lets an interpreter predict, recover from a lost detail without losing the thread, and mark structure in the target — which is exactly what the discourse-signalling work in Chapter 15.4 requires. It is particularly useful for prepared monologue: lectures, presentations, sermons, opening statements.

3. Ingram's semiotic model

Robert Ingram applied semiotic and communication theory to interpreting, framing the interpreter as operating within a system of signs, codes, channels, and noise rather than as a language-conversion device. Its enduring contribution is the reminder that meaning is made from more than language — setting, relationship, gesture, silence, and shared knowledge all carry signal, and "noise" in the channel includes environmental, cultural, and psychological interference as well as literal audio problems.

4. Cognitive load and the Effort Model

Daniel Gile's Effort Model, developed for conference interpreting, is the most useful single framework for understanding why interpreters fail in predictable ways. It treats simultaneous interpreting as the concurrent operation of several effort systems that compete for limited processing capacity:

• Listening and analysis effort — receiving and comprehending the source.
• Production effort — generating the target.
• Memory effort — holding material between reception and production.
• Coordination effort — allocating capacity among the others.

Interpreting works only while total demand stays below available capacity. When any one effort spikes — an unfamiliar term, a fast speaker, degraded audio, a personally distressing passage — the system goes into saturation, and errors appear not only at the spike but for several seconds afterward, while the interpreter is still recovering.

This model explains, in one framework, a great deal of what the blueprint requires:

5. Cokely's finding on lag, revisited

Cokely's miscue research found that interpreters working with a longer lag produced fewer miscues than those working with a very short lag. In effort-model terms this is straightforward: a longer lag buys completed analysis, reducing the prediction errors that produce substitutions and intrusions. In Seleskovitch's terms it allows deverbalization to occur before production begins. Two different traditions converge on the same practical instruction.

6. Using models on the exam

Case-study items rarely ask you to name a model. They ask you to choose the action a competent practitioner would take — and the models are the reasoning that gets you there:

• Demand-Control Schema gives you the vocabulary for identifying what the situation requires and what options you hold.
• Cokely tells you which miscue you are producing and what causes it.
• Colonomos and Seleskovitch tell you that a form-tracking rendering means comprehension has not completed.
• Gish tells you to recover structure when you have lost a detail.
• Effort/cognitive-load tells you which control will actually reduce load: preparation, a team, consecutive mode, a longer lag, or declining.

The consistent through-line is that the correct exam answer is almost always the one that reduces load before the encounter rather than heroically absorbing it during.