I don't feel much for the director of engineering, sitting directly across from me, either. He shows too many PowerPoint slides (never trust a director of engineering who can operate PowerPoint without irritation) that, once I filter all the idealization out of them, basically explain that they tried to hire a few decent programmers to write a generic utility layer so that they could then use a bunch of small teams of cheap junior programmers and disposable contractors to simultaneously write all the bits of their software that actually do things. I'd object to this on moral grounds, even if it had worked, but it hasn't. Their software has turned out exactly as it had to, given what they keep referring to, in apparent seriousness, as their methodology; no two features interoperate, even when they claim to be part of the same work process, and no individual feature solves any hard problem. I think the director of engineering thinks their company is a tragic victim of bad financial-market timing. I think they haven't accomplished enough to rate tragedy.

The other two guys they brought are their technical architects. The architect on the left, at the other end of the table, keeps leaping up to explain the programming intricacies of things the director of engineering has described in broad principle. He has a thick enough accent that I often miss the most important nouns in his sentences, but I get the meta-points quite clearly, which are, on the positive side, that the director of engineering may not know how this thing actually functions, but somebody does, and on the negative side, that what seem like simplistic features, when you look at the software's user interface, are simplistic when you look at the details of the underlying implementation, as well.

There are half a dozen people from my company in the room, too, and over the course of the first six hours of this meeting we have developed a fairly simple response pattern. Their product manager says what problem a feature purports to address, and one of our non-technical executives compliments them on understanding that solving that problem would be worthwhile. Their director of engineering then describes how they attempted to solve it. There follows a dubious pause, and then one of us takes a turn at trying to find a polite way to ask exactly what the hell he's talking about. The architect on the left supplies the details. Then, because our product has all these same features, and I've designed six versions of it, I think of the first three difficult issues we had to confront in building this particular component, and ask if I've understood correctly that they haven't dealt with them.

The other architect, the one on the right, in the middle of the table, has one role in this sequence. After I ask my bullying questions, while the product manager is looking expectantly at the director of engineering, and the director of engineering and the architect on the left are trying to improvise a fable about how this particular fatal flaw could one day, in a future that isn't going to happen, be remedied, the other architect smiles ruefully, half to himself, and quietly confirms that they're missing whatever I thought they were missing.

And it's the precise self-directed ruefulness of the other architect's smile that finally pulls empathy through the mesh of my inquisition and judgments. I have, in a way, been in his place. I have participated in three major software development projects in what passes for a ten-year career as a software designer. The first one was Lotus Agenda, the software for which the term "personal information manager" was invented, only to be subsequently co-opted for glitzy daytimer substitutes. Lotus hadn't run out of money and stopped paying us, of course, which is what these four people whose company I'm condemning are facing, but we did have to argue for our continued existence as a project. We lost. Agenda was too hard to explain, and thus too hard to sell. Lotus canceled it and acquired a cutesy appointment-scheduling program in which the trash can icon burped up a desultory flicker of flames whenever you dropped something on it. I remember the sense of betrayal, kept vivid by the fact that to this day nobody has produced a satisfactory replacement for Agenda. But I remember the ruefulness, too. As much as I resented the Lotus executives' refusal to understand what made Agenda powerful and unique, they were right about their problem. They needed an easy-to-use appointment-scheduling program for the Lotus application suite, and the Agenda team's game attempts (in the version before my move from technical support to development) to dress Agenda up as one were clever, but fairly unconvincing and perhaps even a little grotesque. The new version we were designing would have been cool, but not profitable.

My second project was the Interchange Online Network, a complete proprietary online service (actually a platform for the production of multiple complete proprietary online services) commissioned by the computer-magazine publisher Ziff-Davis, and later sold to AT&T. Interchange was far, far too ambitious, a 4.0 product-design attempting to pretzel itself into a 1.0 release, and it thus took a really long time to construct. In the middle of this really long time the web came along, and once we finally had an operative platform for the production of proprietary online services, nobody wanted such things any more. The subsequent spectacular demise of Interchange, the organization, was much more complicated, protracted and palpably unpleasant than the mercifully quick death of Agenda, but I had the same mixture of emotions about it. On one hand, many of the hard problems we tried to solve in Interchange, about how people interact with virtual communities and combat data overload and information contextlessness, plague the web to this day. On the other hand, by the time we got the thing going it had the approximate clattering grace of Mrs. Tweedy's chicken pie machine. It was a wonder most of it functioned at all, and sadly predictable how frequently, easily and messily the rest of it broke. AT&T wasn't running out of money, either, so I still wasn't asked to pursue a dream on unpaid faith, but I had to watch Interchange die, knowing both what it once might have become, and that that future stopped being connected to our present a long time before the doors closed.

There's a lull in the conversation, a little after three o'clock, in this meeting that is over before it's done. Let it be a moment of silence, in anticipation, for a dying company whose name I think I signed something saying I wouldn't tell you. In fact, it doesn't much matter who they are, or were. They could be anybody. This issue-date is my five-year anniversary at my current company, and I suppose we're marginally more knight than damsel, but it's by no means clear yet whether we'll be successful. I own a lot of stock that is currently worth something in only the most abstract economic sense. I may still get the chance to sit in some other company's conference room and answer for what will come to seem like mistakes. Failing businesses are hardly my generation's invention, of course, but technology companies do fail in some new ways. More intriguingly, perhaps, their failures affect different people. I don't know the other architect across the table very well, but in lieu of contrary information I assume he is just like me. In the old world, he and I were probably never going to have this experience. We aren't entrepreneurs, we weren't going to open stores or restaurants, and the old industries in which we might have ended up lived and died on longer time-scales. I don't know what I was going to be, if the modern venture-capital-backed internet-underpinned software industry hadn't materialized, but I doubt it was anything in which I'd get, just by showing up at the right moment, to be employee number five of what would relatively quickly turn into a two-hundred-plus-person company whose entire business is based on a computer program whose functional specifications I wrote. Maybe I'm oversimplifying the Old Economy, but it seems like dreams used to require much more initial personal investment and risk. What have I risked, to participate in this experiment? Nothing, really. I wasn't involved in the initial product concept or the startup funding, which were the only speculative parts. I quit my previous job on a Tuesday, and started this one the next day. I rolled my 401k over, put the card for the new HMO in my wallet and threw out the old one, and continued on with a pretty easy life. All that stock may never be worth anything, but even without it I've probably (statistically speaking) been paid more than you. It's not my company, by any stretch of the imagination. I didn't found it, I don't own it, I don't control it. I haven't even always been sure I like it. But in the course of working for it for five years, I've invested a surprising amount of emotional energy into it. It now matters to me what happens to it. This is what I could see in the other architect's eyes, even as he acknowledged each of his product's defects. Maybe his software was no good yet, and maybe it never would have become good, but for a while, at least, he had believed in it. He had gone to work, as I do on the good days, thinking that he was a valuable part of an effort to improve the world, in however small or obscure a way. I think there are a lot of us who have had this feeling, in technology companies, and otherwise might not have. And this is the price we pay. A heady thrill has been democratized (if you can apply "democratized" to an industry with such high social barriers to entry), but so too has an agony of defeat. Our parents rarely experienced either end of these arcs. So what do we learn from these sped-up filmstrips? What has tumbling through these spirals done to us?

Tonight, it has just made me sad. After those four people finally left, those of us from my company went around the room, and I voted to let their company die. They had nothing we needed. I wish I'd wanted to save them. In a kinder universe, somehow, they'd get to keep working. Our product wasn't that great in version 1.0, either. Maybe their version 6.0 would have been terrific. But then again, it's not like the mine is closing. They'll all get new jobs as soon as they want them. Part of the twist of that rueful smile may have been the other offers he had waiting. But they'll start their next jobs as different people, maybe wiser in some ways but also more cynical, and quicker to panic, and on balance probably not improved. The people who set up these failures, the CEOs and the VCs, are a tiny minority compared to the rest of us, buffeted by their schemes. So tonight, I want my stereo to make slow, quiet, implacable sounds. I want elegy and requiem, uncorrupted faith, human resolve seeping out of the joints of the machine. I want temporary asylum in a simpler world. I want a sympathetic sigh.

And what I get is beautiful and perfectly haunted. It's been three years since Lucinda Williams' Car Wheels on a Gravel Road, which is easily long enough for it to have settled into a role, in my memory, that isn't exactly the one it really had when I was listening to it more often. I've let it merge with Cheri Knight's The Northeast Kingdom, into a composite epitome of taut, country-, folk- and roots-inflected rock Americana, very much part of the same meandering, extended continuum as the Magnetic Fields, Patty Griffin, Wilco, Steve Earle, Maria McKee, Soul Asylum, Mecca Normal and eventually Sleater-Kinney. A whole bunch of new roots-related albums I'd been waiting for came out approximately at once, a few weeks ago, and encountering them as a set exacerbated my standing tendency to let my expectations impose roles on new records. Often my expectations merely anticipate my reactions, and do no real harm to the process, but in this case they resulted in a mess of confused inattention that took me disconcertingly long to untangle. Untangling confusion sounds like progress, but when I was confused, I had a tall pile of records I thought I liked, and clarity leaves me with a very short one. I wanted 3 of Hearts to be less sickly. I wanted Trisha Yearwood to sound more like Beth Nielsen Chapman. I wanted Nanci Griffith and Mary Chapin Carpenter's albums to be sketchier, Cowboy Junkies' to be less grim. I'm still making up my mind on Chris Whitley, Melissa Etheridge and Stuart Adamson's Nashville project The Raphaels.

The critical mass, however, led to my listening to several of these albums more times than I would have if I'd realized how little I liked them right away. On first listen, Lucinda Williams' Essence bored me thoroughly. I couldn't hear any rock songs on it, which I indignantly truncated, mentally, to "I don't hear any songs on it". I was even a little angry, imagining that this was Lucinda's petulant response to the acclaim for (and accompanying backlash against) Car Wheels on a Gravel Road, deliberately making a record without its qualities. But I kept at it, and around the fourth or fifth time it occurred to me to think of this as a successor not to Lucinda's Car Wheels on a Gravel Road, but to Emmylou Harris' Wrecking Ball. We've already had a successor to Wrecking Ball, of course, Emmylou's own Red Dirt Girl, but Essence shows me a way in which there can be room for two. Red Dirt Girl preserved the atmosphere of Wrecking Ball, for me, but not the effect; Essence recreates the effect without the atmosphere, spare and intimate production as a setting for ethereal, luminous, composure-shattering performances. Eight different musicians play a dozen parts on "Lonely Girls", according to the credits, but almost all of them defer carefully to a terse guitar hook and Lucinda's distracted repetitions of every phrase in the lyrics. She sings "Steal Your Love", over co-producer Charlie Sexton's humming guitar textures and Jim Lauderdale's muted harmony vocals, in a drawl that reminds me of Steve Earle. The first really unmistakable Wrecking Ball moment is "I Envy the Wind", both in its psalm-like lyrics and in Lucinda's elegantly weary delivery, the ends of words fading into her vibrato. "Blue" could be a remedial folk lesson for Jewel, especially if I think Lucinda is talking about the Joni Mitchell album as much as melancholy. "Out of Touch" is one of the few songs with tentative rock intentions, Jim Keltner supplying a casual thump and Sexton adding chirping, vaguely Knopfler-esque guitar and a raspy backing vocal. "Are You Down" is even more Dire-Straits-like, musically, pinched guitar riffs and whirring Hammond, but Lucinda's muttered vocal seems closer in spirit to PJ Harvey or Cat Power. "Reason to Cry" flirts with Patsy Cline, although Patsy's vocal mannerisms don't really work right coming out of Lucinda's throat. "Get Right With God" is a twangy gospel-blues howl, but "Bus to Baton Rouge" is a reverent homecoming lullaby, like an empty-house rewrite of Beth Nielsen Chapman's "Years", which would probably depress me were it not for Lauderdale and Joy Lynn White joining in on a quavery three-part chorus harmony that reminds me of the McGarrigle sisters, allowing the song to evoke family for me even when they're conspicuously absent from the lyrics. "Broken Butterflies", the finale, sounds like it's picking its way through abandoned ruins, but the combination of quiet strings, guitar ambience and harmony vocals in the outro conjures up Wrecking Ball again.

But the centerpiece of the album for me, obediently taking my cue from Lucinda's choice of title, is "Essence" itself. Keltner leans into his drums again, and Williams, Sexton, Bo Ramsey and Ryan Adams all contribute guitars, but instead of serving as tension release, for me the choruses of this song just keep building it, Lucinda and the Jayhawks' Gary Louris singing a methodical litany of anticipation: "I am waiting here for more, / I am waiting by your door, / I am waiting on your back steps, / I am waiting in my car, / I am waiting at this bar, / I am waiting for your essence." It's not much of a text, just written down and skimmed through at reading speed, but in song they draw out each line, leaving me time to contemplate it individually. Waiting by a door is generic, but waiting on back steps implies intense intimacy, a voluminous back story, and arguably a complete cultural milieu. Waiting in a car has its own extensive network of connotations, and simply saying "my car" instead of "the car" is the difference between one life wondering if it will ever merge with the other one and a merged life wondering if it shouldn't have. All of which is only context for what she's really waiting for. A lot of us, probably, are waiting for the people we're with to overcome their fears and share something important with us. We offer to help each other become ourselves, and then have to stand back and wait for each other to admit that there are other selves we're trying to become. We invite souls, waiving the normal rules of civil circumspection. We are painfully, terminally reluctant to burden each other with the mass of our dreams. We sit around a conference table for eight hours, and nobody ever says the most essential thing. Surely nobody really cares about the companies or the products. Most of the things they dreamed of making their product do, ours already does, and most of things we do they probably would have gotten to eventually. What we care about, I want to think, is the belief that we're working towards something we think the world needs, and that whatever tiny steps we take today, however strategically inconsequential they seem, are part of some assembly of truth. We want these companies to become our dreams, whether they're suited for this role or not. And that, too, I see in a corner of that rueful smile, a trace of martyr's nobility. For a moment I can't answer what usually seems to me like an easy question: do we resist corporatism more effectively by infecting it with our ideals, or should we protect our ideals from dilution by keeping them meticulously separate from the compromises our employments insistently demand? Mind you, even when I can answer this question in an instant, I answer it both ways at different times. Some days it's obvious to me that a company is only a sum of individuals, and a sum of sufficiently principled individuals will make a principled company. Some days I fear "sufficiently" is a euphemism for infinite regression, and humans just can't think clearly in groups larger than eleven. And so we oscillate, I guess, between melodrama and cynicism, between frantic gadgetry and mordant Luddism, between judging each other and extending an uncritical hand. We wait, on our own back steps, for our own essence, hoping some of the things we've done will turn out to have been true to it.