Monthly Archives: April 2014

Tragedy of the Commas

Posted on by Standard 1

LegaleseA fittingly long-winded post is afoot for a notoriously long-winded subject. It doesn’t take the due diligence of a top-10 law firm to conclude that just about everyone and their mother hates legalese.

Now don’t get me wrong. The desire to feel competent is an innately human part of mental wellbeing, and language plays a significant role in that aspirational endeavor. There’s a significant grounding in the notion that vocabulary isn’t simply a marker of intelligence, but rather a means of acquiring intelligence as well. Nuanced vocabulary provides a tool for grasping and differentiating the increasingly nuanced concepts seen in complex topics. Besides, I get my rocks off from erudite jargon as much as the next person when it just so happens to suit my fancy and expertise. Ooh baby, nothing potentiates my action potential quite like the perfusion of knowledge that comes from a good peer-reviewed publication. Patch clamps? Kinky.

But abstruse, field-specific language is usually just that: constrained to a certain academic or professional specialty, and not extrapolated to apply to the public at large. Legal wording is perhaps unique in this regard. Encountered on a near-daily basis whether we realize it or not, it’s not only perplexing to the majority of people that it applies to but also binding in ways that the most recent manuscript from Cell is not. If you don’t fully absorb the nuanced ANCOVA variable interactions of a longitudinal cohort study, it probably isn’t going to ruin your day. However, if you failed to fully digest the Terms of Service outlined meticulously by AT&T, your next iPhone upgrade might just become an iHeadache replete with surcharges you didn’t even know existed. Data roaming abroad? Good thing you read the fine print (you did, didn’t you?) because it could set you back $6000… or even more.

The Sparknotes summary is that the florid language endemic to legalese is somewhat warranted by the intricate subject matter that it entails. However, more often than not, it is generally resented, misunderstood, and largely ignored by those that are affected most by its endless chains of commas, footnotes, subordinate clauses, and abstruse terminology. Contracts written between equal business parties are fair game, but for consumer-facing agreements, there’s significantly unequal footing. Why must we “promulgate” policies when we could simply “declare” them or “officially make them known?” Quite the curious oxymoron we have on our hands, given that “legalese” sounds so much like “legal ease.” Too bad the scholars of the world beat me to that punchline.

So the majority of people hate legalese. I know I hate it, or at least grow weary of it. This guy, who is currently both a judge and a Columbia Law professor, hates it. Antonin Scalia has his own doubts about it, and you might just recognize him as a muthafuggin’ US Supreme Court Justice. But more importantly, your own brain hates it.

Why?

Vocabulary is only the very tip of a frigid and unloving legal iceberg: it’s the easiest to pinpoint, but there’s much more shifting beneath the surface to rock our cognitive boats. Everything from the price tag of the products we sign terms of service for, to length of the documents themselves, all the way to the general readability and even the typographic conventions of legal prose tend to stack the attention span and cognitive processing capacity of the individual against their own interests.

So let’s start with an easy one: time. Fully reading — or even skimming — the legal documents we encounter on a regular basis takes a TON of time. There was a bit of stir over the past year when researchers published a study that pegged the average time for consumers to fully read all of the legal material they encounter at 250 hours per year, which totals 31.5 work days for those of you keeping score at home. If you are employed in the US, you might notice that that is about twice as much as the average paid vacation days per year in the private sector, or infinitely percent more time than is required by law (none, the answer is none). Hope you’re ready for a sabbatical.

Reading speed for the averaged adult population is generally considered between 250 to 300 words per minute (the former being used for the research analysis), and that doesn’t even assume 100% comprehension. Given the inherent complexity of legal document content and low intrinsic interest most people have regarding their Flappy Birds download rights, it’s not a stretch to say that those 250 hours might require several dozen more to achieve a decent level of understanding.

Perhaps you think you’ll improve your lot by taking up speed reading? Not so fast, Eyestrain McLightning. While there are currently apps that are doing a dandy job of leveraging technology to assist faster reading techniques via reduction in time spent on lateral eye movement (visual saccade), their potential doesn’t shine as brightly as the smartphone screens on which they reside. The general consensus about speed reading is that with increased speed comes inevitable reduced comprehension of content. A noteworthy pitfall for legal reading, thick with its recursion and subordinate clauses tangled as mangrove roots, which already has a low general comprehension. At normal speeds.

Besides, the public at large rarely finds legal content engrossing. Individual interest and rapt attention can make anything easier to read, at any speed. But terms of service aren’t exactly the gripping type of content that readers consume ravenously for pleasure. No one is going to try to argue that contracts are on par with 50 Shades of Grey; that is, unless your idea of BDSM is getting hog-tied, whipped, and screwed by your cable company. Hot.

I’m not here to judge you or break Rule 34 of the Internet.

Just what is it that makes the 50 Shades trilogy more “readable” than the average terms and conditions? For one, many of the legal agreements that average consumers now frequently encounter on a daily basis are for free services. Fiction we purchase and read is a paid service unto itself, something we exchange money for and expect pleasure from accordingly. We’re invested. With an economy that is becoming ever more data-centric, free shit is a common tactic to get people to unwittingly sign away personal data.

And for humans, FREE is a very peculiar thing. It’s not just college students creeping about the descending margins of twilight, grabbing pizza from club meetings they don’t belong to or fighting sweaty hordes to snag one of the remaining XXL t-shirts their own mother would be embarrassed to sleep in. For anyone who’s ever been to a tech trade show or convention, it’s easy to become inured to the absurdity of grown men and women, dressed in business finery, grabbing up free tchotchkies with the free bags and free totes that they were given specifically to collect their free crap with. And once back home, they overwhelmingly come to their senses, tossing most of it in the trash, to become “free” once again from the unintended clutter of their busy lives.

In short, we’re irrational. Our perception of value does NOT have a purely linear relationship with price. Let’s say we’re selling two items, side-by-side, in which one item is clearly higher quality and more desirable. Assuming that the respective prices offered are discounted below their standing market norms, most individuals will choose the higher-quality product, realizing they are getting a relative luxury for an unusual bargain. However, once prices are linearly shifted down so that the lower-quality item is free, the “purchase” pattern is flipped. Suddenly people are disproportionately gobbling up the free item, more or less blind to fact that the higher quality item is still a better “deal” in the classical economics sense. Put a $2.99 high-quality app next to a related $0.99 one, and many will choose the more expensive option for reasons of performance. Now shift the prices to $1.99 vs. $0.00, and it’s like the more expensive option doesn’t even exist. Amidst our trappings of a free app culture, the lack of monetary investment makes the item more desirable whilst simultaneously decreasing our cognitive commitment to it.

We don’t believe we have as much at stake when we read the legal documents for Gmail as we do when we read the lease for a new car. Thus, we click “I Agree” and move on with our lives, ravenously searching for the next free app to tickle our fancy.

Complexity in writing is another factor, albeit a contentious one. For the most part, we are still abjectly terrible at quantifying our own communicative means, despite the algorithmic nature of society’s most celebrated advancements. However, basic metrics do exist. The most commonly used stat, the Flesch-Kinkaid readability test, was originally developed by the US Navy as a way to classify and standardize readability levels for manuals and training materials. It’s still used widely by the Department of Defense, and also has applications in education. It all boils down to number of words per sentence and number of syllables per word; in a nutshell, lower scores indicate harder reading material. A score of 30 or lower is best understood by a university graduate upward, and publications like Time usually have a score around 50. And if you want to feel perpetually demoralized by your own lack of writing prowess, there’s even a simple way to set up scoring in Word documents. Sure, it’s basic, but a starting benchmark.

Here’s where the fun starts. Most people would assume legal documents would tank on this test in terms of readability score, and most people would be undeniably right. Pennsylvania was the first state to mandate that many insurance policies had to be written to the ninth grade level, or roughly a score in the 60s range. Hilariously, the “Insurance Plain Language Act” itself has a score of 36.8 and grade level of 12.8, but perhaps we’re nitpicking since it isn’t a consumer-facing document. What about other legal items?

I grabbed a business partner contract from a certain litigious “big” and “blue” technology company that I had lying around, and plugged it in. Score of 34.1 with a grade level of 14. Fair enough, given that it’s a business agreement between business parties which are assumed to have business-y legal teams replete with educational trajectories that didn’t fall flat at sophomore year of college. But what about some consumer agreements?

Hmmm.

Again, Flesch-Kincaid scoring is basic, and for the most part, it assumes that big words are overall harder to understand. Not groundbreaking, exactly. But it also has a sort of implied assumption that the reader has the innate capability of understanding the general topics being addressed, with a missed vocabulary word as a minor hiccup. In the legal arena, though, is quite a stretch. One missed word might mean an entirely missed broad and recurring concept. While a grade 12 level reader might have a clear understanding of the five-syllable juggernaut “liability” in general usage, they might still be far from understanding its specific usage and ramifications in a legal context. Most terms and agreements on the worldwide webernets also rely heavily on hyperlinking to connect to more policies and explanations, resulting in an “expanded” content that could equal hundreds of pages were one to attempt tackling it all. Bad enough if it’s in your native language, yes. On the other hand, some tech giants have gotten in trouble recently for sending users to content that wasn’t in the same language as the original terms. Ruh-roh.

Indeed, the jury’s still out — so to speak — on what an appropriate readability metric would be for legal material. Some researchers have recently taken task with finding one, and quickly realized that traditional grade-level scoring metrics did little to predict actual readability of legislation. By using machine learning to compare linguistic characteristics of a balanced corpus of all English writing versus a sprawling sample of legalese, they discovered what everyone knew all along: it really IS harder to understand, with sentence structure highly divergent from other written English conventions.

Well that’s no fun. An online machine-learning assessment for legal readability isn’t yet available to the public, and if it were, it would likely entail downloading yet another app with yet another terms and conditions agreement to slog through. You want a tool now. Hence, I provide a tentative formula — research pending — of my own: (n + 6 = G). Insert the the original Flesch-Kincaid grade level of the document as “n,” and add six years for adjusted grade level “G…” three for the time it takes to complete law school, and three more for the time and patience to thoroughly read the document and all off its attendant hyperlinks.

But let’s go out on a limb and assume you’ve decided to give that gnarly contract the ol’ college try. The battle’s not over yet. You’ve gotten fairly deep into the text, doing your best to remain attentive, but just as you get to the most important and binding parts, something seems amiss. You feel uncomfortable, disoriented, squirming ever so slightly in your chair. And then it dawns on you:

“WHY IS MY LEGAL DOCUMENT YELLING AT ME?!?”

Typographic trickery is at hand. Full capitalization of words isn’t just confined to the shady underworld of internet trolls, but rather is used in legal writing to fulfill the requirement of “conspicuousness” of important clauses. To be fair, it exists for the common good: it makes it harder to bury critical information in monolithic blocks of normal-looking paragraphs, such that a reasonable person would not miss the information. However, the Uniform Commercial Code, which dictates the conspicuousness requirement, only mentions capitalization as one of many options for fulfilling this need (see #10). But hey, color ink is expensive, so full capitalization has become a go-to standard for making important legal junk stand out. But it makes your brain hurt.

No good deed goes unpunished. Paradoxically, the legal community has settled on the fulfillment of “conspicuousness” that most reduces the average reader’s ability to comprehend the very same material being made conspicuous.

All-caps writing is quite well known to decrease readability, to the extent that there is very little current research is even available on the subject. It’s more or less considered a closed case, although interest in the subject has undergone a renaissance within the user experience and user interface (UX/UI) development industry. Much of the work on readability and typography was pioneered by Miles Tinker beginning in the late 1920s, and because we still read and write with many of the same conventions even in the digital world, his findings have remained the authority on what makes some type more readable than others. In some of Tinker’s original research, he noted that all-caps writing, on average, resulted in a nearly 12% reduction in reading speed for adults.

Over the years, there have been several theories on how humans recognize and process words; the faster you can recognize a word as a unit with its associated meaning, the faster you can comprehend text. And for a long time, word shape was implicated as a major factor: with recognition of a particular word’s outline aiding in quicker recognition of the word and its subsequent definition. Even though that simplistic explanation has fallen out of favor (we actually recognize letters and familiar letter sequences within the word unit), the argument against full capitalization still has wind in its sails. Using all capital letters fills space much more homogeneously and creates a clean rectangular edge along the top and bottom of the word, which can incidentally make it harder to recognize those familiar letter combinations that act as heuristic “cheats” for quicker cognition. Less recognition causes more frequent fixations of the eye and more stilted progression of visual saccades, resulting in a slower pace.

Many now acknowledge that the reduced readability of all-caps text is due to a practice effect, being that the average person simply has less experience reading fully-capitalized material. Speed and comprehension can be increased for all manner of bizarre text presentations, including mirror-images, and full capitalization is no exception. If we want legal document capitalization to not slow us down, we ought to simply practice reading it on a regular basis. Perhaps include it in curriculum… hell, in standardized testing. Let us print eBooks and tabloids and textbooks and job applications in all-caps! Let us shout ALL-CAPS from purple mountains’ majesty to echo across fruited plains! Let every caps-lock button be broken and sticky, not for April Fools’ revelry, but for all eternity!

Surely, more exposure and practice would negate the effect of slower comprehension seen with all-caps in a legal context. Too bad familiarity and desensitization would also negate the entire legal reason that it exists: to be objectively jarring and “conspicuous.”

But maybe we’re dissecting legalese from an unfairly Anglocentric perspective. Lest we forget, there are a very limited number of native languages which don’t have recursion, or even cardinal numbers, which so seamlessly permit all those marathon run-on sentences and infinite enumerated clauses. Needless to say, those wireless contracts would likely be easier to understand, given the statistically naught chance that you are fluent in Pirahán or that you get 4G coverage in the remote Amazon.

So, yes. Most legal language is out to screw you in one way or another. And many of the cognitive tricks it employs to do so are incidental rather than purely malicious. It’s up to you to decide exactly what you’re willing to exchange for that next shiny thing. Skim those terms a little closer next time, and just be careful out there.

Wait… you’re not reading this on a free RSS app, are you? Jesus.