Hacker News Reader

Comment by thwarted | Hacker News Reader

I teach contract drafting to third-year law students. It's hard work to take a complex if-then-else statement and render it in plain English.

Is there a reason it needs to be in (harder to write and harder to understand) "plain English" rather than a "complex" series of if-then-else statements? Even dumb computers can understand if-then-else statements.

Or is the "plain English" more valuable because it leaves things open to interpretation after the fact?

View on HN

I draft contracts for a living (IAAL). The goal is always to draft a contract so that any reasonably competent judge or juror could understand the parties' intent. A lot of lawyers (most?) are terrible at this. They say things like "in the event that" instead of "if". They pepper their sentences with meaningless crap like "any and all" or "unless the parties otherwise agree" or "notwithstanding anything to the contrary elsewhere in this agreement" – all of which are utterly superfluous.

The problem is rarely that the lawyer is invoking legalistic concepts. Sure, you might find a few references to statutory laws here and there in a commercial contract, but 99% of a typical commercial contract should be a simple statement of what the parties expect from each other.

Also, non-lawyers tend to get confused by concepts like indemnities, warranties, and limits on liability. These things are dead simple in reality, but lawyers have a nasty habit of dressing them up in coded language. The emperor has no clothes. There are a few places where "magic words" are required by law, esp. around intellectual property rights, disclaimer of seller's warranties, etc. These are the exception, not the rule.

It can get very difficult to clearly express business terms, but it's JUST LIKE WRITING CODE! Case statements, if-then-else, etc. If more lawyers approached contracts like code, contracts would be better. The problem is that contracts only have to "parse" in court. Runtime for contracts is a breakdown in the relationship, and no one believes that the relationship will break down until it's too late. Those of us who approach the contract as something that needs to parse before the shit hits the fan have a different problem: everyone thinks we're overdoing it because they refuse to consider the downside potential.

I could go on forever about this topic.

Since I said IAAL, I must say this is informational only, isn't legal advice, and I don't represent the reader as their lawyer.

Perhaps "Test Driven" or "Test First" contracting will become hip in a few years?

Not necessarily actually testing (as in, let's go to court) but mapping out "given this scenario, x, y and z happen".

Having a contract validation test would be a worthy exercise for the party who didn't write the contract.

This sounds like a great service opportunity for startup/employee contract lawyers... if contracts are like code, then a parser (or maybe legal code pretty-printer) could easily allow a skilled professional to sift through code. Even easier if the contract is standard for a large company.

Any lawyers here care to pick my idea apart?

My firm is working on this, together with other like-minded law firms. Indeed, we are lawyers who write code (gasp!).

> mapping out "given this scenario, x, y and z happen".

I've used scenario tables in some contracts. Each row is a scenario. For each scenario, there are columns for Plan A, Plan B, and Plan C. (Some column entries for a given scenario might be blank.)

There are a couple of made-up examples in a blog posting I did a few years back -- scroll down to "Situation tables" at http://www.ontechnologylaw.com/contract-simplification/.

Interested in additional clients?

// I have to ask here, as you have no contact info in profile. Reach me at my last name on Google's webmail service.

Meta note: It's not clear that your last name is Terretta but that's what I'm assuming. Also, saying "the email system most of us use" on your profile is really confusing.

It just says "Contact me with my name," which is clearly "Terreta." And I assume he means aol.com.

:-D

Of course. I'll send you an email. Meanwhile, my firm's site is here: http://yusonirvine.com/

FWIW, you've got a glaring typo (to me) on your firm's web page:

s/Our job is articulate and defend those positions/Our job is to articulate and defend those positions/

Normally I wouldn't mention it, but since we're talking about both legal contracts and precise prose here...

Excellent site, incidentally. I can mostly tell what you do, which is remarkable for a legal firm's site.

@thwarted, one of the things commonly taught in contract-drafting classes is to break down dense verbiage into (i) subparagraphs, or at the very least, (ii) numbered subdivisions -- like this sentence. See also the IF: ... THEN: ... example in my posting above.

Believe me, "plain English" is greatly desired by just about everyone, not least to head off later accusations of intentional obfuscation. But it takes time (which means legal fees), and to be honest, not everybody is good at it.

I guess my point is that that legalese "plain English" is harder to understand than if-then-else statements. And if it's harder to write to boot, then why bother? Even your example:

IF: It rains at least one inch today but not more than two inches; AND: It doesn't rain tomorrow; THEN: You will turn on the sprinkler system tomorrow; AND: You will not do so the day after.

is ambiguous and difficult to parse (I'm assuming you were using this an example not just for the caps and punctuation, but that this kind of ordering of statements is what is commonly used, even though it was a contrived example). At least two issues I see are:

   - the terminology is based on today but requires knowledge about tomorrow.
     This requires keeping more state to evaluate if the conditions are
     being met for a longer time.
     With this wording, it almost seems like it's setting me up to fail to
     remember to turn on the sprinkler today.
     This would be better worded as about today and having knowledge about
     yesterday.

   - The grouping of the last AND: isn't obvious as to if it's in the body
     of the THEN: or an alternative/conjunction for the entire IF:

The intent would be a lot clearer as something like this (in some kind of mock-pseudo-code):

   if ( (no rain-today) and 
        (rain-yesterday between 1 and 2 inches) and
        (sprinker-not-on-yesterday)      
      ) then {
      you will turn on the sprinker
   }

(at least, that's what I think your intent is, but I'm not quite sure since the goal is still somewhat impenetrable) But even this could be better written with more abstraction, perhaps by defining what it means for the lawn to be sufficiently watered:

   you will turn on the sprinker if (last time lawn received sufficient
              watering was before yesterday)
   sufficient watering is defined as ((sprinkler was on yesterday) or 
              (less than 2 inches of rain occurred yesterday))

(but, really, I'm not sure that matches your intent either).

@thwarted, it'd be great if we could use pseudo-code. (One of my colleagues once proposed using flow-charting.)

Unfortunately, many, many lawyers (and clients) are allergic to contract forms that don't look "traditional." I can say with great confidence that the typical reaction to a pseudo-code contract would be "WTF is this?"

So it's momentum and fear of change (which plays into a fear of no-one-will-need-a-lawyer-to-decode-this, perhaps). Makes sense, but obviously not ideal. What can we do to change this, if anything?

So clearly you have to compile it into legal English. I don't see a conflict here.

A lot of the ambiguity in contracts today has to do with structure, and not so much content. This may seem counter-intuitive, but consider some examples:

1. Lawyers have a bad habit of using "inline definitions" in contracts. That means that in the middle of a long sentence, they'll throw in a parenthesis such as ("Defined Term"). Now, any coder will immediately see that the scope of the "variable" Defined Term is ambiguous without a clear statement of assignment or equivalence. This is a structural issue. The lawyer instead should have put in the contract's glossary: "Defined Term means..."

2. Lawyers tend to use "or" with imprecision. That's why you see many "and/or" in contracts. They either need to use better logic operators, or be precise about logical OR vs inclusive OR.

3. Lawyers get sloppy with timeframes. "Within 30 days of..." is a common formulation in a contract. Do you think the drafter means 30 days before or after? Probably not both. Stuff like this is just sloppy structurally.

4. Lawyers screw up grammar. Commas are really important. Say I list off three conditions: You will do X if (a) thing that might happen, (b) thing that might happen, and (c) thing that might happen with reference to some other thing. Notice the "with reference to some other thing" at the end? If that is preceded by a comma, some courts will apply it to all of (a) through (c). Otherwise, it might only apply to (c). Stuff like that happens all the time.

Now, sometimes ambiguity is OK, or even a good thing. Every question has its own time for an answer, and that time may not necessarily be in the contract. It's important to be pragmatic in a business setting.

Since I said IAAL above, I'm including the standard ethics disclaimer: this is informational only, not intended as legal advice, and I don't represent the reader as legal counsel.

While I can appreciate being pragmatic in a business setting, I find this to be mildly offensive as someone who writes code that, if it isn't unambiguous and isn't explicit, will not do what I want or will crash. Wanting code to operate properly is pragmatic, otherwise you're just wasting your time. Why is leaving contracts ambiguous and open to interpretation pragmatic?

Now, I can see that it's not very pragmatic to quibble over wording/structure in a contract up front, that can just end up wasting time. This is tantamount to purposely writing pseudo-code into your .c file and expecting gcc to do something useful with it -- but programmers don't do that, (the good ones, perhaps those 10x more productive ones) try to write code the first time that the compiler will accept. It seems like it would be even more pragmatic, from a business standpoint, to be more precise in the wording and structure on the first pass and avoid (even the small) risk of there being debate over the interpretation later on. The only reason "being pragmatic" comes up is because it seems to be the norm to gloss over a bunch of stuff and purposely make it ambiguous (considering your 4 examples) rather than being, ahem, explicitly explicit.

I take your point, and it's definitely a fair one. Ideally, we would anticipate and iron out all disputes up front. In software development, that's exactly what we try to do!

In contracts, though, there are cases where a company will live with ambiguity because it has done an assessment of (1) the likelihood of a dispute, (2) its leverage vs the other party, and (3) its ability to prevail on the merits in the event of a dispute.

A contract is usually, although not always, a compromise between two or more parties with at least some divergent interests. In reaching a satisfactory compromise, sometimes you need to prioritize the parties' disagreements and move onward. That's why I mention the concession to ambiguity – because it just happens that way.

And guess what? There's a parallel in software development. Whether it's shipping dates, lack of resources, skills, whatever, software development is also often a compromise. We all know that stuff gets swept under the rug because it's an obscure edge case, or it only affects 0.x percent of the userbase, etc., etc.

My point and yours aren't mutually exclusive; I just wanted to acknowledge that sometimes reality intervenes and makes great things good enough.

Excellent list of the tradeoffs. Thanks.

IIRC Gary Reback points out that, ultimately, legal writing is writing that people are paid to read.

For better or worse, lawyers don't write to entertain and enlighten a general audience...

Reply on news.ycombinator.com