Up: The Passivator

[Related]   «^»   «T»

Regarding the Passivator

By Paul Ford

A response to a response.

Mark Liberman recently published a criticism of the Passivator, which takes me to task in several ways. After composing a long reply, I can't find an email address on his website (a few hours later: I see it was there all along). As his criticism is interesting, and worth considering, and I have said before that I think dialogue on the web should be open and frank at the risk of being impolite, I figured I should eat my own dog food and publish my response here. (To make sense of the below, read Mark's piece first.)

Mark,

Thanks for the critical feedback. My responses:

Though The Passivator is billed as a "passive verb and adverb flagger", it just flags certain strings of characters -- final "-ly" for alleged adverbs, forms of "to be" for alleged passives.

Absolutely true. It's a bookmarklet that flags a regexp, and is often incorrect in what it flags. I like it because it forces me to look at my prose in different ways and become more conscious of the decisions I make while writing, but that's entirely different, to me, than "advice." At its utmost, the Passivator is something that makes the prose ugly and puts some distance between the writer and his or her creation.

Unfortunately, while I made the distinction between "guidance" and "uglifying" when I wrote the bookmarklet, I didn't make it when I wrote about the bookmarklet, and many now see the Passivator as some sort of writing pal. I should have put something in big bright lights at the top of the Passivator piece that says "FOR ENTERTAINMENT PURPOSES ONLY--DOES NOT REPLACE REAL GRAMMAR KNOWLEDGE." I will probably do this, but the damage is done.

I'll limit myself to one small comment. Ford suggests that the sentence "The cat was tired" should be replaced -- because it "should take responsibility for itself" -- by "The cat, sleepy, rubbed David's ankles and mewled -- and was ignored, her desires lost in the gap of language." The proposed replacement is certainly more self-consciously writerly, as well as nearly six times longer. But didn't Ford notice that it also introduces an actual instance of the dreaded passive, "... was ignored ..."?

To defend myself, I was hammering away to make a point, and those sentences are overwrought as a result (and were truly fun to write, for some reason I don't fully understand--perhaps because I knew that they would tweak people). And a little more--yes, I did leave passive constructions in there, and several others throughout the piece (even though they flag bright yellow with the Passivator), because they worked fine and didn't require editing. The point I wanted to make in doing so was that to-be form verbs aren't always evil, and I can see now that this was way, way, way too subtle a way to make that point, especially if someone is hostile to the entire idea, and those who aren't hostile didn't notice. Alack. I also took care to quote a grammar guide: "Don't trust the grammar-checking programs in word-processing software. Many grammar checkers flag all passive constructions, but you may want to keep some that are flagged. Trust your judgement, or ask another human being for their opinion about which sentence sounds best." I referred to this advice as "wise." I get the idea from your blog that you feel I'm an anti-passive zealot, and I think that's based on a misreading on your part.

In my own world, the Passivator fits in with a long line of small text perturbers -- experiments in random-word-and-name generation, long-anagram-making, writing in one syllable words, and so forth that I love to do. The people who've been reading me for years, I think, tend to get it when I'm off in that world of textual perturbation and having fun, but of course when I wrote the Passivator piece I was, if not directly giving advice, implying that I know something. That is the opposite of my personal motto, which is "I have much to learn." (See http://www.ftrain.com/PaulFord.html.). Alas, I should have followed my own advice more carefully, and since no one seems to get what I'm up to on this one, aside from folks who've known me for years, the problem clearly rests with the author.

He takes a bad idea, misunderstands it, applies it earnestly and systematically in a visually attractive form, and then rationalizes its failures as features. Is this what future Semantic Web applications will be like?

That's a fair criticism of the Passivator, but the Passivator has nothing to do with the Semantic Web. The Passivator is based on regexps parsing text nodes in a DOM tree, and has no "semantics" whatsoever--apples and oranges. And, God knows, my role and approach to the Semantic Web represents one of thousands of interfaces and approaches. Dismiss my work, absolutely, but it's an error to call what I do "the Semantic Web. It's like calling the New York Times "the media." There are thousands of people doing all manner of things with the SemWeb, and Harper's is a small drop in a large bucket.

As for your criticisms of Harper's, I can certainly see them. Hopefully some of the work I'm doing in trying to make the content database open and queryable through a simple interface will help people understand what's different about the site. Then again, if they read the site and enjoy the timelines, and can eventually find things they like reading, without ever thinking about what's underneath, that's MUCH better. Honestly, I never expected anyone to have a conversion experience from Harper's alone, which is primitive at best, comprised of static pages, and based on a very few lines of code.

I wonder if your expectations for the Semantic Web--my part of it--aren't too high. Harper's is a big annotated index, which to me is truly fun and exciting, because I love book indices, but God knows that sort of love is not for everyone. If it seemed like the site would deliver more, and you were disappointed, I'm sorry.

See, my whole point is that the Semantic Web, at least for now, should just be about smarter links, and doing things with them, not about modeling reality or trying to keep true to some tech-standards dogma. But other SemWeb folk, and folks like Clay Shirky, keep promoting or criticizing, respectively, the "modeling all reality exactly" goal, which most people working in the field don't seem to see as a goal at all. Personally, *I've* yet to have a conversion experience with the Semantic Web, and the parts of the framework I use are all found in 30-year-old hypertext theory. I wanted timelines that were automatically culled from narrative text [example] and some content semantics, and the Semantic Web was a useful way to get there. When I go to a website and it can tell me ten things I want to read, and all ten of them genuinely meet my interests, I'll be converted. Anything else is just clever database programming--possibly useful, worth considering for certain content management problems, but not miraculous. Which was my point--but again, I guess I botched making it.

In any case, thanks for the criticism. While I think your assessment of me as someone who regularly "takes a bad idea, misunderstands it, applies it earnestly and systematically in a visually attractive form, and then rationalizes its failures as features" is totally wrong and based on misinterpretation regarding the Passivator, and overexpectations regarding Harper's, I can see where you're coming from and acknowledge that I can often be less than clear about what I'm trying to do. I can't expect that I've changed your mind regarding any part of your criticisms, but I'm glad to have an opportunity to think these things through.

Respectfully,

Paul Ford