Rethinking the Future of Law

This is a wiki: Why are you putting text URLs in footnotes?? Link to things, please, in the usual web way.

A short essay cannot have a long introduction. You've spent a paragraph and you haven't told us what this is about yet. You need to put an idea up front in a way that pulls the reader in, and puts her or him securely on the rails that lead where you are going.

Why that's a "transposition" is hard to see from your text. It just means that instead of a world full of printed documents, typed documents, phone message slips, answering machine tapes, floppy disks and video cassettes, we have bitstreams. If that's a "transposition," you need to explain what transposition is and why this is one.

Be that as it may, us juris consultus

It's not clear why you'd want to use obscure Latin here instead of the simple English "lawyers." But if you're going to do it, getting the grammar of both languages wrong (it's "we lawyers" or "we jurisconsults," not "us," while "juris consultus" is singular rather than plural) doesn't help.

lack the know-how, let alone the experience, to confront these challenges. Because of the distinctive characteristics of electronically stored information (“ESI”) and old, bad, lawyerly habits at the discovery level, parties are being shunned from the court system and from having access to justice. For example, ESI is voluminous. Every little piece of electronic storage we possess holds the equivalent of thousands, hundreds of thousands, if not millions of document pages.(4)

"It is estimated" means "Here's a number for whose accuracy I take no responsibility." The number is actually nonsense. Whether a "page" means 250 words of ASCII text, 500 words of ASCII text, one fifth of a short Microbrain Worz document, one fortieth of a long Microbrain Worz document, a TIFF format file of a scanned page, a badly compressed PDF file of a scanned page, a well-compressed PDF file of a scanned page, or a DJVU file of the same scanned page would change the estimate based on bytes per page to anywhere from half a million pages down to a thousand pages per gigabyte. Given that the variability of "page" is more than two orders of magnitude, giving a number makes no sense. Your argument, moreover, is phrased misleadingly, because the nominal storage capacity of objects is irrelevant. Most people's personal filesystems average 90% empty, and after you finish excluding unauthorized copies of multimedia culture and peoples' porn collections, almost no filesystem held in a personal computer contains anything at all, to a first statistical approximation.

And the problem lies not in their magnitude, but in sorting through them to find the few, which are relevant to a legal claim.(5)

But this argument is patently incorrect in the inference it urges the reader to draw. Filtering analog documents printed on paper is difficult: a human being has to read each one. Filtering digital information is easy: computers can do it. Almost all that 81% spam is thrown away before it is delivered to any human being, for example, which is why email is still a very effective medium of communication even with an apparently unpromising signal to noise ratio. Similarly, the effort to use advertising on the web, which has also massively increased the noise to signal ratio of web traffic, makes no difference to me once I install and ad blocking proxy or browser extension, which silently removes the ads from everything before I see them. Your text tells the reader that the problem of filtering grows at least linearly with data volume, but in fact the discontinuous transition from analog to digital representation of the data reduces filtering cost by more orders of magnitude than the volume has grown. Once everything is digital, filtering cost may again grow linearly with volume, but the marginal cost of bit filtering is almost exactly equal to the marginal cost of bit-moving; the difference is so close to zero that you could spend the rest of the lifetime of the universe trying to make it add up to something you could count.

At the same time, lawyers are being swamped with hundreds of expensive proprietary software from e-discovery vendors purporting to have the holy grail of discovery; purporting to hold the true secret for finding the needle in the haystack, the “smoking gun” of a claim. Whether it is through Boolean search methodologies or concept searching, all of these technologies promise results in exchange for “reasonable”, yet “justified” fees for using their exclusive technology. This, of course, translates into exorbitant costs to clients that arise from the handling of volumes of electronic data sought throughout our revered process of “liberal discovery”.(6)

This last sentence implies that the cost of the software rises with the number of bits it processes. I don't think that's really true, even if you are using unfree software to do your filtering.

Lawyers are also being swamped with hundreds of utterly inexpensive zero-cost high-sophistication tools made of free software for doing all the same things. I've seen very expensive email-searching systems that are less good than mairix, let alone notmuch. I've heard lawyers bitching about their difficulty in searching email archives of a few tens of thousands of messages produced in discovery, but I personally carry more than 250,000 of my own messages around with me in my laptop, reindexing the larger archive from which it is drawn—containing more than 1.85 million messages sent me over the years, which is a lot more email than you're going to get in a lifetime—twice a day on two different backup servers, and searching that quarter-million message archive dozens of times a day at absolutely no cost. I built a system last year that could discriminate a subjectively-judged property of patent text searching tens of thousands of patents a minute based on the PDFs of all issued US patents available from Google, using commodity hardware and free software. Making and running the system cost SFLC nothing except a few days' time of one technician and one lawyer.

You are correct, of course, that technological incompetence costs lawyers and clients billions of dollars a year that are paid to slime by morons at the expense of clients who are fools. But the problem lies not in the difficulties of dealing with electronic data in discovery.

But those companies also deal with the most conservatively-organized of the hour-sellers, to whom they pay orders of magnitude more than they pay Axiom. Citing peoples' self-advertisements on their corporate web fronts for the truth of the facts contained is ludicrous, and I can't imagine why you're doing it.

It's not clear to me as a reader, at this point, what the point of the essay is. You've discussed two vaguely-connected phenomena, but you've shown no overarching connection and stated no larger theme. At this point, unless I am professionally required to keeping reading, I as a reader am gone, and you've lost your chance.

Partnership implosion rates are non-zero even in flush times, however, so individual examples are not very probative. Many of the firm failures or near-deaths since 2008 I actually know something about had as much to do with the partnerships' investment decisions as with their practice economics, so I'm dubious about the attempt to tie a short-term detail (the effect of technological incompetence on the economics of discovery practices in large-scale litigation) to the longer-term changes in political economy. I think you're confusing weather with climate.

Maybe. But also maybe not. If people who know what they are doing can use cheap hardware and free software to attain better information control practices at near-zero cost than the idiots who run big law firms are obtaining using millions of dollars, then maybe electronic discovery under conditions of infoglut is a terrain on which we can equalize, rather than one in which the previous disparities between rich and poor litigants will get worse. The piece you cite is another example of bullshit claiming to be scholarship because it's in a law review. Forget what some people who don't know what they're talking about have to say and do your own thinking. Two possible stories about what will happen, each plausible, apparently contradictory. How do you decide what your own view of the situation should be based on?

A new approach to resolving cases is required. And panaceas such as the “initial disclosures”/“meet and confer” provisions, to “tailor” discovery, have already proven unsuccessful. Although there might not be a single approach to embetter the system, absent truthful cooperation among counsel, clients will continue to complain about fees and e-discovery costs will continue to bar parties from the courthouse.

Not a word.

Certainly, many will want to go back to home, to the old business model. But, let them stay home. We don’t want their help.

You have not shown that "e-discovery costs ... bar parties from the courthouse." You have shown that wasteful spending occurs. But the purpose of most litigation is waste. Litigation is a prime example of Thorstein Veblen's point about the centrality of waste in the actual behavior of economic actors. Those who litigate on behalf of poor people don't have to engage in conspicuous vicarious consumption and labor. Their purpose is not to show how rich and puissant their clients are: their purpose is to achieve justice. If technology will allow them to do that more cheaply, technology is opening the courthouse. If the powerful and wealthy continue to use bad expensive technology and ruin themselves doing so, the courthouse door is being swung wider open by their foolishness, and the poor and powerless are getting even more equalization than they might otherwise.