ACM Interactions

Echoing the artist René Magritte’s "Ceci n'est pas une pipe" ("This is not a pipe"), this is not a (we)blog. It's (we)bwords. It won’t log anything. There will be no responses to events or developments. The only event is the posting itself, with each aimed at making words reappear. My intention is to focus on the timeless, and leave the briefly timely to others. The timeless is forever timely, so stay tuned. Don't run off, because you can't hide. You can't even outrun words.

Words are inescapable, 'bwords or otherwise. There can be no science, research, or scholarship without words, but they are more often seen as part of the problem, than part of the solution. Hence, technical writing seeks to write out writing. Objectivity, distance, and avoidance of the literary combine in the hope that not one word will ever stand out. Words, and the experience of reading them, should all disappear, leaving only unsullied knowledge. I want to put the words back into the mouths of all who find them unpalatable. 

Popular discourses, including those around folk science, tend to bite the words that form them. This mistrust of words has a long history. Renaissance humanists were scornful of medieval schoolmen, the scholastics who dominated the first centuries of European university education with their dialectical disputations, careful conceptual analysis, and deft drawing of distinctions. Scholastics were falsely accused of asking, "How many angels can dance on the head of a pin?" Similarly, in ancient Greece, philosophy and rhetoric teachers who charged for education in argument and disputation, the Sophists, were the origin of sophistry and sophism as pejoratives. In more recent centuries, the British have used the "gift of the gab" as a similar pejorative, while physicists worldwide rein in legitimate differences of position with "stop arguing and start calculating."

Words then are not to be trusted. Nor need they be when facts can speak for themselves. Renaissance scorn of scholasticism heralded the empiricism of 17th-century science, with its preference for evidence over argument. However, reasoning within science remained strongly influenced by rationalist traditions, with naive empiricism remaining an extreme position. Even as the hypothetico-deductive model of science became established in the 19th century, the correct balance of evidence and reasoning remained a matter of contention: Mill and Whewell thus disagreed on the required extent of reasoning in advance of any claim of confirmatory evidence. Whewell could not see how facts could ever speak for themselves. Words must speak on their behalf in advance of their arrival.

The boldest and clearest attempt to define and defend scientific method was known as logical positivism. Developed in the mid-20th century, truth was not a matter of naive empiricism or, mathematics apart, pure reasoning, but the appropriate combination of both. Freddie Ayer, one of the main proponents of logical positivism, admitted later in his life that "I suppose the most important [defect]...was that nearly all of it was false." Both the logic and the positivism had been punctured in the 1940s. The former bubble was burst by Quine, who showed that claimed legitimate steps in verbal argument ("analytic statements") were all circular in nature. The latter bubble was burst by Popper, whose black swan exemplified the impossibility of ever securing adequate samples for irrefutable inductions. Both Quine and Popper were analytical philosophers decades before the wild men and women of postmodern relativism. The bubbles were burst from within, by philosophers with commitments to rational science. However, misconceptions endure among barrack room philosophers of science, hardly surprising since very few of those who peddle uncritical Scientism have engaged with the history, philosophy, or sociology of science. In short, few advocates of naive Scientism are scientific about science, having never engaged with the relevant (historical) evidence. No facts can speak for themselves in an audience that can't listen.

Listening isn't always enlightening. Words can be used in bad faith, with poor logic, and without evidence of critical reflection. They can be non-committally vague and without consequence, a clear position, or a basis for probing their truth. However, all such sins are avoidable. Words and argument however are not. Without words, there are no arguments. Without arguments, there are no design rationales. Without design rationales, there are no convincing designs, experimental designs included. The experiments of HCI and other research are all stillborn without a well-reasoned and plausible experimental design. Too often however, critical reflection and candour is reserved for a closing "discussion" section, with its plea bargaining attempts to confess to some intrinsic flaws in the hope that others will be forgiven. Whewell, John Stuart Mill's opponent on 19th-century philosophy of science would have seen through this. The best science is the result of extensive prior reasoning, often transformed by outstanding creative insights. The resulting studies need no discussion section to bail them out. Rigorous prior critical analyses, and the insights that these often release, take care of that.

Even so, there is typically someone, sometimes sometwo, on CHI panels who will casually dismiss another’s critical conceptual analysis with "I don't see where that will get us," or "So everything's too hard and nothing's possible." In response to the former, this may unwisely assume that we already are somewhere. If you are currently nowhere of any worth, then rigorous conceptual engagement may be the only way to find the first effective steps towards anywhere. In response to the latter, this may uncritically assume that our current blinkers and shortcuts will let us make real progress. At worst, such positions degenerate into philistinism.

Words are not here to be tamed. They are here to be followed. We must take them seriously. Rigorous research does not start with a hypothesis or conjecture. It starts well before that. Still less does rigorous research start with the gathering and analysis of evidence. Everything has to be thought through beforehand and afterwards, and that is impossible without words. 

Taking words seriously is a serious business. We don't do it enough in HCI and interaction design. In every disciplinary endeavour, vocabularies shape conceptual repertoires. Concepts provide interests and direct attention. Not only do facts not speak for themselves, they cannot even appear without concepts to categorise and form them. A discipline's critical concepts are more than technical terms. Terms do not start and end with definitions. Instead, they are located in historical discourses that continue within contemporary research cultures. Getting to thoroughly know the key words in any discipline involves more than learning a definition. Every key word has stories to tell, and these stories may be inconsistent. Some stories bring out new sides of words that then reappear to our consciousness. Through such scrutiny, everyday words, work words, tame words, and taken-for-granted words can suddenly reappear as exotic, uncooperative, unruly, and demanding.

I'd planned to end my first 'bwords with some HCI examples of conceptual naivety, narrowness, confusion, confounding, and abuse, but I've decided otherwise. Instead, I will leave these to future 'bwords, and restate my main positions:

Words get a bad press [sic], but this often originates in another one-sided position, such as favouring the free over the commercial, facts over concepts, results over reasoning, or analyses of data over critical conceptual analyses.

Words are all we have, primarily and ultimately: Argument does not reduce to positivist logical inference, and evidence does not reduce to positivist sensory data. Research and design practices start with words, in their motivations, rationales and imperatives, and they end with words, in their claims, arguments, lessons, insights, and conclusions. Data only comes to life in a well-tuned chorus of concepts. Data powers science, but it doesn't drive it. You need words to do that. Words are the hands on the steering wheel. We need to drive carefully. Taking care with words is taking care with science. Equations apart, theories are 100 percent words. The task of science is to form and test theories. Words are our friends here, not our enemies.

Words are thus worthy of serious attention: Anti-sophistry is misguided, and words are undervalued in scientifically oriented research. Key words are lenses that direct and focus our attention. The wrong lenses in the wrong place blur. The right lenses in the right place expose vital details that would otherwise be overlooked.

In future 'bwords, I will consider key words and phrases for HCI: how they have been defined, how they have been used, how they have been misused, where they came from, and what comes with them, irrespective of whether one knows any of this. So, please feel free to request or suggest anything that’s not on my current list for consideration, which is: affordance, all, design, ensure/guarantee, natural, needs and wants, optimise, requirements, user, validate, works. 

Gilbert Cockton