Tuesday, October 23, 2012

Avoiding irrelevance and dilution: Construal-level theory, the endowment effect, and the art of omission

Deciding what to omit (omission) is one of three fundamental writing skills—besides fluency and cohesiveness—supporting the two major Writing Virtues: Clarity and Concision. Although omission is a sophisticated skill not acquirable through panaceas, it is unique among the three fundamental skills because a single roadblock causes most of the congestion. The roadblock is the writer’s innate aversion to deletion; the aversion derives from a universal cognitive bias called loss aversion, meaning we’d rather maintain the status quo than bet a significant amount on the flip of a coin. (See D. Kahneman, Thinking, fast and slow (2011).) The most dramatic expression of loss aversion is the endowment effect: owners will sell property only at a much higher price than they would pay to acquire it. Loss aversion explains an impressive part of wordy or irrelevant writing because it makes adding matter easier than deleting it.

Construal-level theory is a theory about decision and judgment that explains loss aversion and teaches us how to avoid it in writing. Construal-level theory deals with the biases the distinction between practice and theory introduces into our thinking. When our objectives are immediate, the information available rich, and time bountiful, we analyze in a way of thinking called near-mode, which uses high-grain, concrete concepts and attends to incidental features. When our objectives are long-term and the information or time scant, we analyze in a way of thinking called far-mode, which uses low-grain, abstract concepts and focuses on the essential.

Construal-level theory furnishes an explanation of loss aversion and the endowment effect. (D. Kahneman, supra.) Consider a standard example of the endowment effect: a holder of concert tickets costing $50, the most the concert-goer would have paid, refuses to sell for $300. More usually, we value property about twice as much just because we happen to own it already. Construal-level theory explains loss aversion by the tendency to give greater importance to the near than the far. Analogously, we over-value what we’ve written because it’s near, and we’re loathe to part with it.

Construal-level theory has unearthed another source of our reluctance to cut inferior matter: the audience reads in far-mode, which is global, but legal writers often compose it in near-mode, which is sequential. The consequence is that the audience averages the quality of the documents’ parts, whereas the writer is apt to add their quality, meaning that, for the audience, subpar arguments detract from overall quality but, to the writer, they may seem to increase the quality. (K. Weaver et al., The presenter's paradox (Oct. 2012) 39 Journal of Consumer Research 445 [Hat Tip: Overcoming Bias].)

Construal-level theory provides insights to help writers overcome the biases implicated in writing in near-mode for an audience reading in far-mode. Writing systems involving different roles for the author, such as the roles of writer and editor, serve to vary the author’s mode. Specifically for brief writing, Bryan Garner has advanced a more elaborate system of roles, which are distinctively near and far. (B. Garner, The winning brief (1999) at p. 3) The chart below displays the Flowers roles, their typical activities, and the mode mainly engaged.

Flowers-paradigm role
Typical activities
Mode from construal-level theory
Brainstorming, “Deep thought,” background research
Outlining, planning, detailed research
Primary writing
Editing, proofreading

Madman is far because it encourages intuition, a far-mode product. (See G. Gigerenzer, Gut feelings: The Intelligence of the unconscious (2008).) Architect is near-mode because it accentuates logical relationships, which depend heavily on sequencing, a near-mode activity. Carpenter is far-mode because it attempts to make ideas intelligible to others. Judge is near-mode because it involves close reading for error.

The alternation of phases is powerfully effective in engaging both modes without causing the mutual interference to which they are prone when combined simultaneously. It is so effective that the modes can be seen to alternate within roles. Although Madman is predominantly far-mode, it includes periods of near-mode activity, such as close reading of selected cases. Judge, although mostly near-mode, may include far-mode phases, such as hearing the document read aloud. Carpenter and Architect usually alternate more than once, because Carpenter excels at abstraction and Architect at sequencing.

Construal-level theory offers far-mode as the remedy for excess. Because of the relationship between the endowment effect and near-mode, cutting excess is performed most effectively in far-mode, and typical problems in legal writing occur when lawyers compose their briefs in near-mode, often because they write their briefs while they read cases closely. The result is not only the absence of the big picture but also an accumulation of excess. To avoid much of this excess, learn to write in far-mode, and master the research in near-mode in the Architect phase.

Some common advice is misguided because it contributes to excess. Writers are often instructed to be Madmen in the Carpenter role, but although both Madman and Carpenter are far-mode, Carpenter provides the opportunity to pare down irrelevant matter generated in the Madman phase, and the advice to suspend the critic when doing primary writing sacrifices the main opportunity to trim excess. This shouldn’t be left to the Judge, as editing is a near-mode activity, and among the errors the Judge isn’t good at correcting is excess. The erroneous advice comes from seeing an alternation between writer and critic rather than between near-mode and far-mode.

Both far-mode phases are good for trimming excess, as the writer can take steps to stem excess in the Madman role despite heeding the advice to suspend the critic. Although this advice applies to Madman, not to Carpenter, when applied to Madman it admits critical comments. To take advantage of the Madman to combat excess, treat critical ideas related to scope and breadth just as you would any other ideas. If you’re brainstorming, if you think you’ve come up with an idea of doubtful relevance, you should note that thought alongside the idea itself. Having added exclusion as an idea, you will later be unable to avoid discarding one idea or the other, the marginal thought itself and the imperative to disregard it. Therefore, near-mode’s reluctance to part with sentences will resist rejecting the idea to discard as much as it resists rejecting the idea itself.


  1. I write fiction, and I don't think people read fiction in far mode. I could say the goal is to suck the reader into near mode. But "near" and "far" are too simple--without thinking deeply about it, I reject the notion that every mental activity falls into one of these two opposites. Thought is too complex. Things break down into two opposite but arbitrarily-constructed categories, like Republican & Democrat, when there are game-theoretic reasons for that to happen. This is probably not the case with our own mental activities. I question whether the benefits you find in assigning things to these categories is greater than the information you erase in doing so.

    Writers, and their objectives, are too different for any piece of advice to be good for all writers, except possibly "learn to type faster". Some writers write sparse outlines, and expand, not prune, them in the "carpenter" phase. Some writers, like Pynchon, write primarily for the beauty of individual sentences, and it is impossible for them to do their "madmen" work in far mode, since their creative ideas apply only to individual sentences. Ray Bradbury writes stories start-to-finish without planning, and he's the best in the world at what he does.

    1. Far-mode and near-mode represent a dimension rather than a dichotomy.

      I'm an avid fiction reader, but I don't write it. Fiction can be (relatively) far or near. But writing it is probably always nearer than reading it. (Common advice to fiction writers is not to try to mimic real conversation in dialog because it's boring. Such mimicry is apparently hard for writers to resist.)

      When there used to be book stores, you could find far-mode fiction in the section labeled "Literature." Tolstoy's "War and Peace" is perhaps the best example of far-mode fiction. The reader is disinterested: he isn't apt to identify with the characters, which is part of what "suckers people in."