A modest proposal for the use of metaphors in technical writing
August 22nd, 2010 by ravi

In the 1994 surprise hit Il Postino, the Chilean poet Pablo Neruda is shown impressing upon the town postman and aspiring poet, Mario, the importance of metaphor in poetry. Good poetry is impossible without clever metaphors, but for more mundane writing the device and its close relative, the analogy, are perhaps best dispensed with. For, more often than not, they seem to do more harm than good, when it comes explicating a concept, plan or analysis. Take the following confusions…

A race to the finish?

At the large corporation at which I work, an intelligently crafted presentation that I was witness to included nonetheless a strange use of metaphor: to stress the need for the company to react with agility and speed to the demands of the market, the presenter could have used those very terms (“agility and speed”) with sufficient markup or decoration, in his slide(s). Instead, perhaps for greater effect, he chose to use images to depict the difference between the company today and the one it need become to succeed in the future.

The unfortunate images he chose were those of an elephant (to represent the current lumbering giant of a corporation) and a cheetah (to suggest the corporation remake itself in the image of the fastest recorded land animal). This is unfortunate since his metaphors, rather than sealing his point, argue for quite the opposite considerations! For in the game of Darwinian survival, it is the resilient generalist, the elephant, that has won the race, while the over-specialised cheetah is near extinction (having sacrificed much evolutionary opportunity to achieve its impressive speed).

Sometimes, an analogy or metaphor may turn out to be quite dangerous!

A bitter cup of tea

Some of those opposing the current US administration’s ambitions and programmes have anointed themselves the “Tea Party” — in doing so they cleverly associate themselves with the founding of the nation, and its libertarian roots as they see it. This self-identification paints a picture of citizens struggling for individual rights against an overbearing government that is intent on collecting and putting to waste their hard-earned money. If their argument or association is valid, the analogy they employ is simple and evocative, just the sort of good purpose a metaphor can serve (on the other hand, if their underlying thesis is invalid, this highlights another danger of metaphors: that they can often mask the real issues and obtain agreement via emotional appeal; that danger is addressed in the next section).

In response, their detractors — the defenders of President Obama and the Democratic Party — have launched their own “party”, the “Coffee Party”. But unlike the shrewd label “Tea Party”, this one presents no ideological picture or emotional appeal to its members or the general public (that they wish to influence). If anything it reinforces the derisive image of liberals painted by their opponents, the image of effete coffeehouse pseudo-intellectuals.

A mountain out of a molehill

The mathematics and computer science world has been abuzz lately with the announcement of a proof that P ? NP (a very important unsettled question in computer science / complexity theory). Worry not if you are unaware of P, NP, etc., for they are not critical to the point to be made. The claim of a proof was made to a small number of mathematicians and computer scientists, but was rapidly disseminated through the research community, drawing comments from far and wide. Which led to the concern that should the proof turn out to be wrong (or worse frivolous), then the precious time of some very important mathematicians, who were pulled into the verification effort, would have been wasted irredeemably.

To help explain this loss of time, one of the commenters on a popular blog discussion of the proof offered the analogy of a hiker stranded on a mountaintop. The call for rescue, the analogy goes, results in serious mountain climbers “mount[ing] a brilliant coordinated effort … to get the hiker off the mountain”. The analogy purportedly helps us see how the lack of preparation, as well as other errors of the hiker have wasted valuable resources and time.

At first blush the analogy seems to convey in clear terms the terrible consequences of soliciting expert opinion without prior rigorous (and lengthy) individual effort. However, while the problem may be real, the analogy is in fact deeply flawed. As might be evident, the flaw here is that a call for rescue, imposes an ethical imperative that cannot be ignored, is hardly similar to a proof verification request which can be entirely ignored without guilt or treated with a level of importance of one’s choosing.

Metaphors and analogies are dangerous and misleading beasts. It is, I think, time for the surgeon general to opine on them and alert the public on the menace. Until (s)he does, I propose the following rules of usage:

1. Eschew the metaphor when words suffice. Metaphors should clarify or help the reader’s intuition grasp a complex concept in a simple way. If the concept itself can be expressed in a sentence or two, do you really need the metaphor?

2. Ensure your metaphor is “cashable” (as the philosopher Jerry Fodor writes). Given the metaphor, you (and the reader) should be able to produce the more complex paragraph or two that the metaphor aims to neatly summarise. If there is no underlying explication, the metaphor is mere hand-waving.

3. Tickle the brain, not the heart. Getting people to agree with you by appealing to their sentiments is not always cricket. Aim the metaphor not at imparting a good feeling but at clear thinking. Motivational metaphors are best left to the purveyors of inspirational doodads.

Sign up and support my call for this 3-point prescription to be included in the Microsoft PowerPoint license agreement.

The Hurd affair and HP’s greatness
August 16th, 2010 by ravi

A recent New York Times article (link at the end of this post) addresses the puzzlement that has occupied many, about the flimsy reasoning behind the dismissal of HP CEO Mark Hurd. The article sheds much necessary light on the larger reasons for this action, but what also caught my attention in the piece was this section:

Charles House, a former longtime H.P. engineer who now runs a research program at Stanford University, openly rejoiced when he heard that Mr. Hurd was leaving. “I think the sexual harassment charge was a total red herring,” Mr. House told me. He didn’t care. “I was delighted,” he said.

Mr. House’s brief against Mr. Hurd went well beyond his outsize compensation and penchant for cost-cutting. As Mr. House saw it — indeed, as many H.P. old-timers saw it — Mr. Hurd was systematically destroying what had always made H.P. great. The way H.P. made its numbers, Mr. House said, was not just cutting any old costs, but by “chopping R.&D.,” which had always been sacred at H.P. The research and development budget used to be 9 percent of revenue, Mr. House told me; now it was closer to 2 percent. “In the personal computer group, it is seven-tenths of 1 percent,” he added. “That’s why H.P. had no response to the iPad.”

I am always bemused when I hear this “HP great, but failing to compete now” argument. My questions arise on two fronts:

HP’s greatness and research/innovation

Without doubt HP is a tremendously successful company. It dominates its field, ranks #1 in mark share for certain products, and can boast a few firsts as well. I have no particular beef against the company (despite HP/UX) and I think they were, as claimed on their web site, ahead of their times in some of the human resources practices, such as offering flex time, employee profit sharing and stock options, and employee health insurance. But what is also noticeable on the HP timeline web site is that what has flowed from HP Labs is a steady stream of evolutionary advances, not breakthrough inventions or innovation.

Compare the HP time line to That of AT&T’s Bell Laboratories. Bell Labs is (or was) the home of seven Nobel Laureates, the birthplace of the transistor, the Unix operating system, the C and C++ programming languages, information theory, the laser, as also significant advances in diverse fields such as photonics, astronomy, cryptography and digital music encoding.

My intent here is not to “diss” HP or HP Labs. HP’s ability to productise its advances, and in turn to target productive fields of research, is made apparent by the same comparison with AT&T (Bell Laboratories), and the current status of each: HP continues to be a successful and healthy company while AT&T and its spin-off Lucent Technologies (now Alcatel-Lucent) are languishing with little to show for their pioneering achievements.

HP’s consumer presence

When HP acquired Compaq, the latter had already lost a good bit of its standing as a top consumer computer systems manufacturer. Nimbler outfits like Dell had stolen the mantle from Compaq, and though HP leads the pack today in US and global desktop and laptop sales, I suspect this leadership is a result of corporate and government contracts, rather than consumer purchases (unfortunately I could not find any reliable data online that breaks down computer sales into consumer and corporate/governmental categories).

News and data are available for other forays by HP into consumer devices, and that news doesn’t speak well. For example, in 2007, HP abandoned the digital camera market, having shrunk from a meagre 7% market share in 2006 to 4% in 2007. HP’s acquisition of Palm and it’s possible entry into the mobile communication space and reentry into the tablet space will provide another data point. But even there, with the dominance of Apple and Google’s Android based systems, we can somewhat safely predict that HP would be lucky to gain a double-digit market share.

In summary, (once again) my wish is not to trash talk about HP. A company that has demonstrated some level of decency towards it’s workers is commendable for that action alone. What I would like is for the media and bloggers to give the rest of us some insights into what they mean when they talk about HP’s greatness and innovation. For, it seems to me, HP’s lack of an answer to the iPad has nothing to do with lack of R&D, but likely more to do with the very strategy that makes HP successful: creating or leveraging relationships that facilitate volume sales. That works for netting multi-million dollar orders from the US Government, but, among other things, being in bed with Microsoft and at their mercy, rules out any chance of their developing something like the iPad (by the way, I think the iPad is over-hyped, but that’s another post). What am I missing?

David Foster Wallace on Perl Programmers
August 15th, 2010 by ravi

Okay fine, he is talking about a particular kind of tennis trainee, but it seems worrisomely applicable to those (like me) who are unable to part ways with Perl:

You’ve got the Complacent type, who improves radically until he hits a plateau, and is content with the radical improvement he’s made to get to the plateau, and doesn’t mind staying at the plateau because it’s comfortable and familiar, and he doesn’t worry about getting off it, and pretty soon you find he’s designed a whole game around compensating for the weaknesses and chinks in the armor the given plateau represents in his game, still — his whole game is based on the plateau now. And little by little, guys he used to beat start beating him, locating the chinks of the plateau, and his rank starts to slide, but he’ll say he doesn’t care, he says he’s in it for the love of the game, and he always smiles but there gets to be something sort of tight and hangdog about his smile[.]

But still, how is it possible to easily bid adieu to an unpretentious language that provides a construction as tight as: $x ||= 5;

The Blackberry Apex
August 9th, 2010 by ravi

You know that old saw: “first they ignore you, then they ridicule you, then they fight you, then you win”? It occurred to me that it applies quite well — except for the ‘then you win’ part — when it comes to technology and the attitude of business jocks. Every bit of technology created and embraced by geeks is ridiculed by men in suits as, well, geeky, upto a point where it’s value becomes apparent to them; shortly after which they turn ridiculous in their addiction to it. This second point (of addiction) in the life of that technology, almost always identifiable by the near-hysterical adoption of a totemic device, we refer to as the BlackBerry Apex.

The chart below depicts this finding for one technology: electronic mail or e-mail.

Corollary: not every BlackBerry Apex has a corresponding iPhone Recovery.

Disclaimer: I attempt to kid! Some of my best friends wear suits!

Time to Wave Goodbye
August 4th, 2010 by ravi

And so another great Google experiment comes to an end. CNN reports that Google Wave is on its deathbed:

Google is pulling the plug on Google Wave.

Google intended the messaging program, launched in 2009, to be a near-replacement for e-mail, which it said had grown tired.

But on Wednesday, the company announced that it is shuttering the project by the end of the year because it didn’t have traction with consumers.

“Wave has not seen the user adoption we would have liked,” Urs Hölzle, a Google senior vice president, writes on the company’s official blog.

The real experts in the blogistan will be weighing in shortly with acute analysis on the cause of this sad outcome. But before they do, I figured I would get some shots in.

Google Wave was the one Google product that excited me after Google Maps. Google Earth – cute but what’s the point? Google Checkout – less icky than Paypal but how far can that go? Google Apps – little, though not too late. So on. But Google Wave was an attempt to solve a problem that faces anyone who uses email and messaging in a serious way i.e., anyone who predates the arrival of Blackberry. Keeping track of conversations, action items within conversations, expanding conversations to a larger group, and a host of other needs were the very target of Wave (though it had other goals as well).

Why then the failure?

I can offer no more than agreement with the consensus that it was altogether too complicated and complex a tool — and productive use of it required that one’s collaborators use it as well — but the one deficiency that applies to so many of Google’s products (such as the Gmail and Google Reader web interfaces) was ironically missing in Wave: Google’s arguably poor UI aesthetics. I am not going to once again link to the “41 shades of blue” New York Times article; suffice to say that Wave refreshingly abandoned that Google norm of HTML 1.0 styling, stark colours, and crowded elements co-existing with chunks of empty space. Widgetized boxes with drop shadows, pleasing shades of blue and green, a nifty scrollbar, clearly demarcated buttons and menus, and other sophisticated elements set Wave’s UI apart (let us ignore, out of respect for the dying, the use of Arial over Helvetica).

Google has noted that bits of Wave will be reused in other products. I hope they start with the user interface.

»  Substance: WordPress  »  Style: Ahren Ahimsa