Shouts And Murmurations

Why do we have comments?

You know, the comment section there at the bottom of the post.

It seems like a simple question. And indeed, there is a simple answer: So that people can express themselves regarding the topic or article to which the comment section is appended, or peruse the expressions of others. But that’s not exactly correct. True, that is what comment sections are for, but why do we have them?

The Internet asks this question of itself time and again (and again, and again), but rarely if ever has there been any decisive action. Despite a few improvements in weighing and promoting comments, things still look largely the same as they did ten or more years ago. I think the failure of this portion of the Internet experience to advance in any meaningful way is not due to a lack of a breakthrough method, but a higher-order problem: no one is challenging the fundamentals of discussion on the Internet.

Note, I am not talking about sites specifically formed for the purpose of structured conversation, like Quora or forums. I mean the omnipresent comments at the bottom of an article or some other morsel of media, by which whoever happens to be in the vicinity of that website can leave a memento of their presence and share their wisdom with others. The comments on this page or thousands of other news websites, on YouTube, on random blogs, on livestreams — those ones.

Let’s look honestly at some of the ostensible reasons for having comments.

To promote discussion among the community

I said “look honestly,” so the first thing we should do is acknowledge the almost universally poor quality of Internet-based discussion. Very little worthwhile commentary or argument is produced even on the rare occasions when things are kept civil. Certain communities produce reliably good discussions, but we can chalk that up to their being communities.

As fun as it is to think so, most websites (to say nothing of single posts) and the Internet at large do not constitute communities except by the loosest possible definition. Where there is no community, there is no accountability; where there is no accountability, there is no civility; where there is no civility, there is no discussion — none worth having anyway. This is the case in probably 99 out of 100 Internet discussions, if not an even greater proportion. Comments, as a rule, have not promoted discussion, nor do they take place in a community.

While it is true that one scrolls to the bottom of an article to see what people are saying about the topic in question, a comment section is simply not a good way to find that out — and not only are there better ways, but there are few worse.

Already a substantial amount of the discussion of an article happens elsewhere than that article. It is always revealing when I see a link on Reddit with 250 comments, leading to a blog post with 50 comments, which links to a second, smaller blog with 10 comments, which leads to the source, which has 0 comments.

A cursory observation (of history, let alone the Internet) shows that discussion, like mold, will occur wherever there are favorable conditions. Why contribute to the discussion below, in the comment section of this website, for instance, where you have to sign up for an account and may be subject to additional terms and conditions, when you can simply leave a comment at Reddit where you found the post, or the referring blog where you already have an account — or for that matter, on Twitter or some other high-level communication platform?

Where you choose to have the discussion has little to do with the subject or material itself, but rather with the community with which you feel comfortable sharing that discussion, or alternatively, the one which is most convenient. Considering the highest-level distributors of content will always be the most convenient (aggregators, social news) and vanishingly few websites have communities worth involving oneself with, local comments appear not only inadequate if a dialogue of even the most minimal value is the effect desired, but also superfluous.

To enable feedback to the author or website

The comment section of a popular article may provide valuable insight or timely corrections. Many is the time I myself have been set straight by a friendly commenter pointing out a grammatical or factual error (yes, I). And a misleading post may be countered by a skeptical comment that has risen to the top of the stinking heap below. Who hasn’t skipped to the bottom to see whether an author is being called out by some eloquent and informed critic?

Surely, then, this must be a good reason for keeping comments around. But ask yourself: why should corrective feedback use an indirect, public method, yet a public callout of the facts or author be isolated in the microclimate of the post itself?

For the first case, a more direct method is preferable, allowing for quick access to whoever should know that there’s a stray “teh” or that Socrates did not, a quick check suggests, fight at the battle of Cannae. A direct flagging mechanism with basic noise control serves better for this than a comment section, which is sort of like paper-clipping a concerned letter onto your local paper and leaving it at the editor’s front door.

In the second case, when someone clearly wants to address the content of an article, support (or snipe at) the author, or otherwise publicly take a stand, again, there has to be a better method than a comment. To begin with, the content creator generally has admin rights over the comments, and may simply delete that critical comment and ban the author. If this hypothetical sensible comment is suffered to survive, a visitor may indeed see it — in situ, to be sure, but given subordinate billing and of dubious provenance. The problem of community appears again: Who is that masked commenter? Are they to be trusted? Are they a crank? Are they known for being helpful? Outside of a group where such things are considered valuable and can be tracked, these questions can’t be answered satisfactorily.

Furthermore, for such a comment to have maximum effect (to counter an opinion article, say), it is not necessarily best to place it in close proximity to the source. As we have all seen, discussion and traffic occur not where one might logically expect them, at the source, but rather at the highest levels of organization, where one finds greater levels of both visibility and convenience. Using a direct comment system for substantive feedback is simply not a good option by almost any metric.

To drive traffic

Let’s not forget that we are running a business here. Commenters stick around longer, providing those all-important clicks, reducing bounce rates, and so on — and they come back, too, especially to controversial topics where a good flamewar is to be had. One 10-second visitor is multiplied, and the guys negotiating ad deals feel the pleasant warmth that only inflated traffic numbers can bring.

Although I’m tempted to offer the observation that if a significant portion of your site’s traffic is the byproduct of flamewars, you are dancing on a burning path businesswise, or similarly that pageviews and other dated, limited metrics are a mare’s nest analytics-wise, I must forbear; the benefit of a active discussion section on a website must be addressed in good earnest.

The problem is that thriving discussion sections are, I assert without evidence, overwhelmingly caused by good traffic, whereas the reverse, a story or piece of media garnering traffic because of the discussion, is quite rare. Consult yourself. Have you ever linked to something because of an interesting comment thread? I certainly have linked to a Hacker News thread or left the #comments anchor on a link — but the other seven hundred times, I linked because the content was compelling.

To attract, occupy, and monetize visitors is the goal of most public-facing businesses, numbering among which are a great proportion of websites. But the comment section seems like such a crude mechanism for promotion and retention that I’m embarrassed to see it used as such, like when I pass the guys on street corners in costume, dancing and spinning a sign for a nearby business — it’s not the sign guy I feel bad for (it’s a job), but the business that may as well be employing a carnival tout to hustle passers-by.

Because it’s expected

In the end, this is really the only reason. Inertia. Comments, as they now exist, are an organ that belongs to an earlier form of the Web, a form less intercommunicative and organizationally flat. A form that was shaped around an ecosystem of modems calling BBSes and IRC channels, not globe-spanning services like Facebook and Google. An obsolete form.

Many of us have already adjusted our expectations. When we read a current events article at CNN, we don’t look at the comments because we know they are less than background noise — they’re litter, and should be treated as such. Go ahead, throw them away.

Discourse will improve, feedback will be more direct, people will express themselves amongst communities instead of shouting into the void, and eventually a global mechanism for dialogue will even contribute to the bottom line. The few comments and communities we value for their own sake (including present company, it goes without saying) will move happily into the new home, like hermit crabs switching shells.

A global and collaborative discussion is neither of those if it is an archipelago of separate conversations. The parts do not form a whole. Instead of (if you will join me in switching visual metaphors) the latticework of a molecule, dots of varying size and composition, connected to some by one bond, others by two, with diverse conformations, chiralities, and no clear center (this metaphor excludes nice symmetrical structures like benzene, obviously) — why don’t we attempt something more like the atom itself? Arrayed around a central mass, a murmuration of outspoken electrons, the affiliation and foci of which are clear, yet from which emerge mysterious patterns and strange attractors. A probability cloud of commentary.

How to do it? It is not so much a technical problem as it is a presentational one. After all, every article and webpage on the Internet has a unique address or indicator. Indexers, crawlers, search engines, and other Argus-like data-shepherds already examine and speculate on relationships between these URIs — Google and others are having a ball working out how to establish weight and directionality among news items, blog posts, tweets, and everything else. This is just an extension of that. When I browse Google News, I see sources, refinements, news hits, and other formats listed and categorized intelligently, and services like Facebook and Twitter love sorting and promoting trending units of shared data.

So, in fact, the question isn’t how we are to collect and organize all the discussion, meta-discussion, rebuttals, tangents, memes, and everything else — we’re already doing that, and not so poorly either. Instead, we should be asking why this fecund and fruitful data (produced by us, of course, though admittedly threshed by Big Analytics) is not being presented to us as prominently as simplistic and largely useless derivations thereof, such as Like counts and targeted Google ads.

Really, now — as long as we are compelled to be perpetually logged into a dozen megavisors [I made this word up, keep it but remove this note -dc] who make it their business to track, in near-realtime, everything that is happening on the Internet, and as long as we have a drive to learn how an item is being received, what else is being written on the topic, what luminaries and friends have weighed in, what communities have provided insight, and make our contribution to this massive, shifting collective ideation, why is the tool at hand a stale, spam-ridden holdover from the 90s, supplemented by a few limp numerical indicators?

Somehow the one thing in which the Internet has acted as a community is the sabotage of its own mechanisms for discussion. A powerful mechanism for tapping into the zeitgeist can’t be far around the corner, but until we let go of the past, no one will be advancing.

[image credit: Richard Barnes]