Skip to main content

comments on "clarity"

My paper on "An Institutional Approach to Computational Social Creativity" has passed the "conditional accept" hurdle that was presented in the first reviews I received from the Seventh International Conference on Computational Creativity:

I think the author has made a good effort to improve his paper. My recommendation is to accept it.

But there are still some problems: 

Nevertheless, there are some points that I believe the author should try to improve for his final version.

My thought in this note is to document a few of the remaining pain-points.  I suspect they're likely to come up in other papers I write, too, at least until I get a lot more practice.  In particular, reviewing some of these issues should be useful preparation for the "Phatics, phaticity, and phatic studies" paper.

Here's how the reviewer continues:
The abstract and the introduction can be better. Please, check the following link:
http://cs.stanford.edu/people/widom/paper-writing.html

The clarity of the document still can be improved. I insist on this point because our main goal as authors is to connect with our readers. So, for instance, see section “4A. Monitoring users”. This section has two main paragraphs; however, they are not properly linked. It is necessary to clearly, maybe it is better to say explicitly, see how the content of the first paragraph relates to the second one. Probably my concern has to do with the fact that the provided information in the text is too compact; so, it might not be that easy for some readers to connect the dots. Please, check the whole document on this respect. 
The link points to a document by Jennifer Widom that seems like a micro version of the book "Writing Science".  It provides some section-by-section tips about technical papers.   This does seem useful.

In connection with the remarks above, I'll extract a couple relevant quotes -- and add my worked exercises.
The Abstract: State the problem, your approach and solution, and the main contributions of the paper. Include little if any background and motivation. Be factual but comprehensive. The material in the abstract should not be repeated later word for word in the paper.  
My current abstract:

Elinor Ostrom’s Nobel Memorial Prize-winning work on “the analysis of economic governance, especially the commons” scaffolds an argument for an institutional approach to computational social creativity. Several Ostrom-inspired “creativity design principles” are explored and exemplified to illustrate the computational and institutional structures that are employed in current and potential computational creativity practice.

I think the reviewer is correct -- the problem, approach, solution, and contributions are relatively buried or hidden in that. Let me try again.

Problem: Modelling the creativity that takes place in social settings presents theoretical challenges: there is relatively little agreement about what "social creativity" means.

Approach: Rhodes's classic "4Ps" framework offers a simple and relevant typology. Elinor Ostrom’s Nobel Memorial Prize-winning work on “the analysis of economic governance, especially the commons” shows how to build a theory of social creativity with more rigour and detail.

Solution: Several Ostrom-inspired “creativity design principles” are advanced and used to explore the computational and institutional structures that relevant to computational creativity research.

Contributions: These principles frame a literature survey that describes the shared concepts that structure the contexts that support creative work. These concepts are connected to the idea of computational "tests" to foreground the connections with standard computing practice.

...So presumably I can just delete the underlined terms and I'll have a new (slightly longer) and better abstract.

Making room for some longer text is a bit of a challenge but I guess I'll just need to delete some text in order to simplify the presentation.

The reviewer comment that the "text is too compact" seems totally accurate.  It's an 8 page paper, so there's only so much I can say.

Perhaps I can use another idea from Widom's outline to fix some of these compressed sections?

Here's how she describes the introduction:
The Introduction: [...]  Here is the Stanford InfoLab's patented five-point structure for Introductions. Unless there's a good argument against it, the Introduction should consist of five paragraphs answering the following five questions:
  1. What is the problem?
  2. Why is it interesting and important?
  3. Why is it hard? (E.g., why do naive approaches fail?)
  4. Why hasn't it been solved before? (Or, what's wrong with previous proposed solutions? How does mine differ?)
  5. What are the key components of my approach and results? Also include any specific limitations.
Then have a final paragraph or subsection: "Summary of Contributions". It should list the major contributions in bullet form, mentioning in which sections they can be found. This material doubles as an outline of the rest of the paper, saving space and eliminating redundancy.

I'm not going to go through the introduction text here, but instead, will try to see if I can apply these ideas to the aforementioned Section 4A -- assuming that Section 4A can itself only be an introduction to the ideas, given the space limitations.

4A. Monitoring users. [Proposed design principle:] “Tests document the interaction of Producers and Place.”

A shared environmental sensor may notice when a given Producer makes some specific kind of change to the Place. A Producer’s own sensors also notice changes in the Place. In the first instance, what a Producer produces is sensory
data. Information capture is thereby more or less localised, and may be shared more or less widely. For Ostrom, monitoring helps deal with conflicts and provides grounds for sanctions. Here, it becomes clear that monitoring is a requirement for any action or interaction whatsoever. Along with sensors, action requires effectors, or (more broadly) transducers.


Traditional unit tests document the success or failure of implemented code relative to a pre-existing standard. A broader notion of testing can also compare the result of any action with the initial concept, e.g., comparing a painted image to an initial altered snapshot, as done by The Painting Fool (Colton and Ventura 2014). A similar idea can extend to “proprioceptive” sensing and judgement about effected actions. Similar judgements about upstream data allow for filtering and the associated “ecocompositional technique [of] constraining” (Keller 2012). 


  1. What is the problem? 
    • The straightforward view suggested by the idea of "monitoring" is to deploy some global functionality that keeps track of the actions of all participating Producers within a Place.
  2. Why is it interesting and important?
    • Sensors are generally deployed along with effectors or (more broadly) transducers that translate the sensory information into action. So, monitoring is important for modelling any action or interaction whatsover.
  3. Why is it hard? (E.g., why do naive approaches fail?)
    • For example, The Painting Fool compares an initial snapshot (sensory data) to the painted image that it generates in response to that snapshot, and judges the quality of its output on that basis.  This example could be extended to theorise "proprioceptive" sensing and judgement about effected actions more broadly.
  4. Why hasn't it been solved before? (Or, what's wrong with previous proposed solutions? How does mine differ?)
    • Filtering upstream data is another simple application of sensors and transducers; for Keller (2012) this is one of a several important "ecocompositional techniques."
  5. What are the key components of my approach and results? Also include any specific limitations. 
    • The general conclusion is that distributing the monitoring function among participants is vital for social creativity.
That's only a little shorter (after removing the underlined text -- 170 words as opposed to the original's 180) but I think it is a lot clearer.

The straightforward view suggested by the idea of "monitoring" is to deploy some global functionality that keeps track of the actions of all participating Producers within a Place.  But this function can be broken up and distributed out among the Producers themselves. In the first instance, what a Producer produces is sensory data. Sensors are generally deployed along with effectors or (more broadly) transducers that translate the sensory information into action. So, monitoring is important for modelling any action or interaction whatsover.  For example, The Painting Fool compares an initial snapshot (sensory data) to the painted image that it generates in response to that snapshot, and judges the quality of its output on that basis.  This example could be extended to theorise "proprioceptive" sensing and judgement about effected actions more broadly. Filtering upstream data is another simple application of sensors and transducers; for Keller (2012) this is one of a several important "ecocompositional techniques." The general conclusion is that distributing the monitoring function among participants is vital for social creativity.

A further comment to connect this with phatic studies


It seems to me that there are several things going on here that would be really useful for us to theorise (and apply!) in our work on phatics.  For instance, why do the five numbered points in Widom's theory of Introductions seem to work rather well (and, apparently, not just for introductions)?  Perhaps it's because these points help to draw attention to the things that really matter in a given text.  Perhaps one of the ways they do that is to filter out extraneous material. 

To be clear, this is much more in line with a Jakobsonian phatics -- and the changes above don't appear to be about small talk.  Indeed, my writing has previously been criticised for being "overly chatty."

However, what the reviewer said above does suggest that clarity connects with "relationship goals":
The clarity of the document still can be improved. I insist on this point because our main goal as authors is to connect with our readers.
And I think this is why a rejected paper, or even just constructive criticism like I've received here, can feel so embarrassing.  In a La Barrean mode, phatic communication is about understanding (cf. Slide 5 in "A schematization of phaticity").

To further illustrate my thought here, I followed up a now familiar intuition and learned that "Clearness" in Greek is σαφήνεια, where the -φήνεια component seems to link with our cluster of "phatic" terms.

So I think one branch of "applied phatic studies" would be devoted to developing clear communication.

Relating this to some ideas in the recent Phatica 3.1. blog post: 

[socialibility] "is behind the possibility of even phatic communication among them"

This reminds me of the familiar writing trope "consider your reader."

Also, if
"links of fellowship" are "consummated only by the breaking of bread and the communion of food" [or, perhaps, among those] "who [have] attended the same classes, read the same books, seen the same entertainments, and [know] the same people"
...then I think the particular kind of fellowship in academic discourse is very much of the latter sort -- and lack of clarity makes that kind of communization difficult or impossible.

In a simpler or "primitive" communication context, the activities of phatic communion also seem to depend on clarity.  An unclear greeting, or rudeness or lack of concreteness surrounding food -- these sorts of things could probably escallate to all-out war.  Malinoswki refers to "the specific feelings which form convivial gregariousness" - the specificity seems almost as important as the gregariousness.

One final thing to say here is that "clear communication" seems to result in "clear text" -- so the ideas here could start to bridge into our upper-right quadrant of topics related to "future phatics."

Comments

Popular posts from this blog

The plot thickens (with Herbert Spencer)

In a paper attempting to outline the conceptual domain of comparative psychology , Herbert Spencer discusses the quality of impulsiveness in relation with human races (bearded and unbearded). Among his "sundry questions of interests" about the relationship between mental energy, evolution, complexity, etc. are the following notes: ( b ) What connection is there between this trait and the social state? Clearly a very explosive nature - such as that of the Bushman - is unfit for social; and, commonly, social union, when by any means established, checks impulsiveness. ( c ) What respective shares in checking impulsiveness are taken by the feelings which the social state fosters - such as the fear of surrounding individuals, the instinct of sociality , the desire to accumulate property, the sympathetic feelings , the sentiment of justice? These, which require a social environment for their development, all of them involve imaginations of consequences more or less distant; and th

Vitruvius Pollio, The origin of the dwelling house

 Chapter 1 of Book II of "Ten Books on Architecture", available from Project Gutenberg .  Sections 1, 2, and 7 (from the Richard Schofield translation published by Penguin rather than the one here) are quoted on pp. 218-219 of Spheres II by Peter Sloterdijk.  Pay particular attention to Section 2. 1. The men of old were born like the wild beasts, in woods, caves, and groves, and lived on savage fare. As time went on, the thickly crowded trees in a certain place, tossed by storms and winds, and rubbing their branches against one another, caught fire, and so the inhabitants of the place were put to flight, being terrified by the furious flame. After it subsided, they drew near, and observing that they were very comfortable standing before the warm fire, they put on logs and, while thus keeping it alive, brought up other people to it, showing them by signs how much comfort they got from it. In that gathering of men, at a time when utterance of sound was purely individual, fro