Technical Audacity And The Phone Book

I often think we — or maybe the people who control our money — lack the audacity to take on really big projects. It is hard to imagine laying the transatlantic cable for the first time today, for example. When I want a good example of this effect, I usually say something like: “Can you imagine going to a boardroom of a major company today and saying, ‘We plan to run wire to every house and business in the world and connect them all together.'” Yet that’s what the phone company did. But it turns out, running copper wire everywhere was only one major challenge for the phone company. The other was printing phone directories. In today’s world, it is easy to imagine a computer system that keeps track of all the phone numbers that can spit out a printed version for duplication. But that’s a relatively recent innovation. How did big city phonebooks work before the advent of the computer?

Turns out, the Saturday Evening Post talked about how it all worked in a 1954 article. We aren’t sure there weren’t some computerized records by 1954, but the whole process was still largely manual. By that year, an estimated 60,000,000 directories went out each year in the United States alone. Some of these were small, but the Chicago directory — not including suburban directories — had over 2,100 pages. In New York City, the solution was to print a separate book for each borough. Even then. the Manhattan book was three inches thick and projected to grow to five inches by 1975.

It’s in the Book

If you aren’t old enough to remember phonebooks, they were printed in tiny small print — usually in a typeface called Bell Gothic — and on very thin paper. So five inches of book adds up to quite a few names. We don’t know how they arrived at the figure, but the article claims the book is 91.91% accurate. That seems like an awfully specific number and it makes us recall the old adage: “88.72% of all statistics are made up on the spot.”

Only nine printing presses in the country were able to handle the printing and binding of thick directories. Another thirty shops could handle smaller ones. But we can only imagine the amount of work that went into collating, updating, and typesetting. The article also talks about something that would be easy today but was a massive undertaking in the precomputer age: directory assistance.

From the article:

Manhattan Island alone has five information exchanges; at each one some 90,000 queries a day are received in automatic rotation by long rows of operators. The girls sit in glass-partitioned booths, surrounded by borough and suburbant directories. Close at hand is the printed daily addendum of 500 or so new numbers, which comes to the information operators about twenty-four hours after the phones have gone into service.


If you’ve never seen a real phone book, the Library of Congress has a whole collection you can view online. Need some plaster work done in Los Angeles in 1940? They’ve got you covered.

For some people, this was their big chance to see their names in print.

Phone books have been slowly dying for a long time. It isn’t surprising that some Canadian students had trouble using one.



It is hard to think of projects quite this ambitious and audacious today. Maybe it is because, with computers and other modern tech, the whole undertaking would be manageable. Sure, we are going back to the moon, but that’s been done and with modern technology, it is less audacious than it was in 1969. We mobilized COVID 19 vaccines in record time, but that was in response to a crisis.

Sure, we spend plenty of money on big projects. Expensive telescopes and supercolliders come to mind. Quantum computers are expensive, but it just isn’t the same scale. Where are projects that truly break new ground on a massive scale? Maybe the hyperloop if we ever build one. It is hard to define why, but telescopes, supercolliders, big computers, and hyperloops don’t seem to have the grand scope of the phone system, the transatlantic cable, or delivering 60 million phone books a year. There are a few examples. The swarms of satellites to provide Internet access feels like a large-scale project. Stringing fiber everywhere there are phone lines isn’t very exotic, but it is a lot to handle.

What do you think? What’s the most audacious engineering project of our time? What’s the most audacious one you’d like to see? A space elevator? An underwater city? A flying airport? Let us know in the comments. Meanwhile, [Mike] thinks we shouldn’t even be using phone numbers anymore. As for big projects, sure they are costly, but the payout can be huge, too.

74 thoughts on “Technical Audacity And The Phone Book

  1. “What’s the most audacious one you’d like to see?”

    A Plan-9-ish, open-source operating system written in Ada that doesn’t force people to use a radically idiosyncratic user interface.

    I’d donate $1 to the folks who step up and do it.

        1. Yeah, well, so do inches, feet, miles, pounds, and gallons, but it don’t make them right.

          Ada (not “ADA” – ADA is the American Dental Association, while Ada is named after a person) is sort of the Esperanto of computer languages. It was developed to reduce the number of langauges that had to be suppored by the Department of Defense, and at least in theory, all software not written in Ada had to have this fact justified before the DoD could buy it.

          And by the way, I don’t think that either the Apollo Guidance Computer or the computer that controlled the Saturn Launch Vehicle was programmed in Ada, since these were embedded systems and therefore exempt.

  2. Wasn’t there a post about Audacity just last week?I

    They say Minitel in France was implemented to do away with phone books. Hard to believe a terminal in every home was cheaper than a phonebook..

    1. Minitel was not just to do away with phone books – in fact I remember my aunt and uncle using their phonebooks while they had minitel – it was a basic information service, in some ways a precursor to how we use internet now. There were several lookup services, weather, phone, stocks, train delays, and rudimentary information websites.
      It was especially advantageous to small businesses, e.g. farmers could do their milk planning and taxes over minitel.
      It was also state-sponsored.

  3. The article makes it sound like there was one organization printing all those phone books at once.

    It wasn’t.

    It was a whole bunch of smaller organizations all doing it in similar ways.

    Yes, Bell was one company.
    No, the central Bell office did not coordinate the printing of all of the phone books

    Each local branch did their own.

    That’s how you take on massive projects – divide and conquer.

    Nobody said “lets wire all the houses in the world together.”

    Somebody said “lets wire a bunch of houses together.”
    Somebody else said “lets wire these bundles of connected houses together.”
    Before long, loads of houses are connected all across the world – and nobody set out to do that.

    Smaller groups set out to do smaller tasks.

    The sum of those smaller tasks is the world spanning telephone and data network we have today – all built by small groups making it easier to talk to other small groups.

    1. Well said. Exactly my thoughts.

      As a business model the telephone system worked at many scales. There were those early adopters that were willing to pay big for connections to a small number of other wealthy users. As it scaled up the cost were shared with more and more users and the cost per user came down, such that more users could afford it.

      By contrast, think of the Panama canal. This was a huge project that cost many lives. Such a project did not have any utility until it was ALL finished. You couldn’t build a small canal for small boats and scale it up gradually to large ships. This is why it takes governments to invest in such projects.

      1. The transcontinental railway was built by the Central Pacific and Union Pacific as cheaply as they could. As the completed railroads earned money the infrastructure was greatly improved. In some cases routes were abandoned and new ones built. The Federal government did provide subsidy by land grants along the newly built railroad. However, this also was an investment by the Federal government because it greatly increased the value of the nearby land retained by the government.

  4. Whenever BT is complaining it’s too expensive to give everyone fibre to the premises I think of the gas network.

    Flammable and explosive gas piped under pressure to nearly every home with near 100% reliability.

    Fibre has got to be easier hasn’t it? Same goes for water and sewers.

    1. Not all homes (in America) have gas piped to them, natural gas is only piped to homes in populous areas. My last house had a propane tank, and a well for water and a septic tank.

      1. The problem with this scenario (your last house) is that the majority of the population has no concept, or understanding, of the virtue and utility–and absolute need of self-reliance; nor is there ANY understanding, at all, of what is truly essential (“…but I absolutely NEED my cellphone. And wi-fi, And…”)
        The only segment of our population (world-wide) which does, in general, truly understand what is really essential for survival is farmers, and others who “live off the land”.

        The only thing missing from your list (which you probably had, to handle power outages) were kerosene lamps (they used to be called ‘hurricane lamps’, and for good reason) and a generator capable of running your well pump.

        Hope you can get your present house UP TO the elegant standards of your last one.
        Best of luck…


        Those ads on TV, for “whole-house generators” really make me laugh. And the timing is impeccable–they started appearing, with a vengeance, immediately after the debacle in Texas this past winter. And they continue to this day…“Never be dependent on your power company again…”, they proclaim. The implication, of course, is that all one has to do is throw money (and GOBS of money, one quickly finds, after investigation) at the advertiser, and ALL your power problems will be magically solved. Of course they will; then you CAN have your cellphone, your wi-fi, your Facebook, and all those other essentials without which you absolutely canNOT survive…

        1. Thats… thats the point of technology and social constructs. To create a better place for the next generation. To allow THEM to make a better place still, etc. There will always be those who have never known hardship, and that is fan-fucking-tastic. I hope my kids don’t need to deal with the hardships I have…

    2. Maybe if we made phone lines explode every time they fail, BT would suddenly discover they could make them reliable.
      But even BT haven’t a scratch on Thames Water: 30% of clean drinking water lost between purification and the consumer. That’s enough to give everyone in Sudan clean drinking water.

  5. The idea of rolling out 7 billion handheld supercomputers sounds pretty audacious to me. Or to sequence the entire human genome in a few short years, and later hours. I also think we’ve gone from “let’s do one really impossiblly big thing” to “let’s do an impossible number of impossibly small things”.

    Also, do not mistake Musk’s overtures into space as a billionaire’s pet project. We are entering an age where entire celestial bodies are for the taking with those with the resources to do so. If you think Jeff Bezos is rich, you haven’t met the toll collector on the highway between Earth and Europa.

    1. Hype, I’ve heard the same BS for 30 plus years. We can barely put men into orbit let alone do a manned Moon landing. despite having far more technological advances than the Apollo and Shuttle engineers could dream of.

      They put men on the Moon with slide rules,

      All Musk has done replicate some Apollo era tech and polished it up. Which isn’t bad but is a testament to how far we had regressed. Prior to Musk, we were relying on the Shuttle which was designed before you were born and probably the last hurrah of the old school NASA and aerospace engineers.

      Then we stagnated for decades, lost boat loads of engineering talent which is almost impossible to replace.

      Manned trips to Mars or asteroid mining is a fantasy with today’s level of space tech,

        1. The main function of the ground crew is to protect the investment, and to double-check all of the hard math, when things go wrong. So you want cheap, trouble-free Earth-Moon and Earth-Mars trips? That’s going to take a while, although not as long as it took the last time someone got serious about space flight. In 1969 it was unthinkable to have rockets automated to the point where the boosters could land on their butts and be used again. Now it’s considered an essential feature of any new launch system, for an acceptable price per kilogram. The whole spaceplane notion, which seemed like an easier task, was great for sci-fi, and for getting congressional funding, but not so great in the real world, because, you know, physics.

          I think it’s crazy that the only reason we have viable space projects today is that in the capitalist world, it is possible for individuals to get so stinking rich, they can “launch” today’s audacious projects. But the corporate world isn’t built for audacity, so that’s where we are.

          1. To the future, the first rockets will seem like taking a flight to Paris, jumping out with a parachute and just letting the 747 crash would to us. Absurdly wasteful, insanely obviously not ideal.

      1. Google should just team up with the USPS, and put a street view camera on every delivery truck. You could probably have damn near 100% of the most populous areas updated every 24 hours, and probably quite a few rural areas as well, seeing as they deliver to just about every damn US address 6 days a week

    1. Yeah, I’d say that’s on the same scale as phone books. But even online satellite maps is a scalable thing, because you can start off with one pixel per 10,000 meters and work your way to ever-higher resolution.

    2. I think the GPS system was pretty audacious. It played a significant role in making Google Maps and Streetview possible, as well as making it useful. The idea of using precise timing signals from satellites to allow anyone anywhere on Earth to be located within a few meters by using a cheap handheld receiver sounded crazy when I first heard of it.

      It still does. The more I learn about the internals of how it works, the more I admire it.

      It’s amazing the number of things that now depend on GPS. Many of them have little or nothing to do with navigation, but rely on GPS for time and frequency distribution.

      1. That’s true. It’s not just the system, satellites had to develop, and electronics had to evolve so a package that could fit in your pocket was good enough.

        Fifty years ago, lots of equipment, and a big antenna.

        But fifty years ago, it was LORAN.

  6. I’d like to see nationwide fiber to the home. Western Electric had all the infrastructure ready to roll, I’ve seen the catalog, so it’s possible, and needs to be rolled out to replace all that abandoned copper. If the ISPs won’t do it (and they’ve been stalling for decades), maybe it’s time for government to step in and subsidise the rural areas. Municipal fiber (also subsidised) where the cable companies are the only providers, because hardline analog infrastructure is a dead end. Fiber everywhere for everyone.

  7. “…What’s the most audacious one you’d like to see? A space elevator? An underwater city? A flying airport? Let us know in the comments.”

    The elimination of hunger and disease everywhere on this planet. Forever.

      1. Who mentioned humans? So – all hunger then, and all disease.

        That would be an interesting scenario.

        Maybe we would all starve to death (and the animals too) as hunger is what drives us to find food?

        No disease to kill the plants, so we would end up with all plants and no animals.

  8. Okay, well, I don’t know much about Rust, but here’s why I think Ada would be a good idea:

    It’s a pretty safe language and has a decades-long history of being used in safety-critical software. It’s equally focused on reliability, maintainability, and efficiency. And in my (limited) experience, it does a very good job at that.

    Ada’s grammar forces the programmer to express problems more explicitly than other languages. The fact that it has a very strict type system, for example, helps to prevent a lot of errors. And GNAT is just ace when it comes to catching compile-time errors in a way that is immediately useful to the programmer.

    Also, Ada is, IMO, considerably more accessible than, say C, which is more like a thin layer over assembly language and thus constantly forces you to “think” like a computer. And while it offers OOP features, Ada is, at its core, a fairly simple procedural language.

    Last but not least, Ada is an ISO standard, and has been for decades; and it seemingly hasn’t suffered the collateral damage of hype.

  9. “We mobilized COVID 19 vaccines in record time, but that was in response to a crisis.”

    Hmmm a genuine hack for which long term impact will only be fully realised by those in the know.

  10. Yes the phone books were printed in Bell Gothic, and in house. They were also published using UNIX on their choice of DEC hardware and an appropriate printer. Funny thing, the K&R book on C were also printed that way, and largely composed using Emacs and then feed through Troff and then done that way.

  11. An “easy” one that would reap world recognition and admiration would be to solve world hunger. That would take a lot of organization, but we have all the technology we need. It’s just a matter of spending the money on it instead of something else. For example, I’d love to build or see built a space elevator. That would take a ton of work and new tech, and keeping it safe from idiots with radical ideologies would be the real feat, but we should instead just make sure we can all eat first. However if the US would stop killing people, we could use the budget to do both at the same time. 700billion is a lot of billions.

    1. If solving world hunger was just a matter of spending the money it would be done already.

      That is literally the equivalent of “couldn’t we all just get along?” No. We can’t. For a lot of reasons that are extremely complicated and hard to solve.

      Phone book logistics aren’t any more audacious than many of our current projects. To think they are probably means a lack of appreciation for how technologically advanced we already were in the past combined with an individualistic mindset. The idea of throwing massive brute force behind large problem sets is a frame of mind. “We” as a species still do it today, but it’s easy not to see it.

  12. “Can you imagine going to a boardroom of a major company today and saying, ‘We plan to run wire to every house and business in the world and connect them all together.’” Yet that’s what the phone company did.

    Not really though. It started out with major companies and trickled down to smaller and smaller subsets over time and over technological advancements. that’s why it worked, they started small and waited for technology to catch up before broadening the scope.

  13. 700 billion handheld computers is relatively trivial in the scope of human endeavour. We already had both computers and worldwide mass production capability, so the “how” was essentially covered. The only remaining question was “what”, and the intersection of demand and cost addressed that for us. From that perspective, the difference between a mobile device and a hair dryer or machine screw is negligible.

  14. The massive change in infrastructure that will come about because of climate change. It’ll effect agriculture, transportation, water distribution & recovery, and energy distribution. All while the weather is trying harder to kill us & threatened societies are calmly and rationally laser focused on those issues.

    Whether that change is a massive collapse or a leap in efficiency remains to be seen, but it’ll be spectacular either way, and will cause some ripples that promise to be pretty bodacious on their own.

    1. “Whether that change is a massive collapse or a leap in efficiency remains to be seen”

      It will be neither. It will be a gradual evolution and adaptation. Some of the changes will be acute responses to medium-sized catastrophes, most will be so gradual as to go unnoticed until somebody looks back decades later.

  15. I’d like to see the same massive amounts of money, engineering talent and perceived sexiness currently thrown at space travel to be given to grid scale energy storage and long distance transmission, one of the only things standing between us and a renewable energy revolution. Hand $10bn to a private company and say “I expect you to build something which can store enough energy to power this entire village for a week”. Large scale storage and transmission would mean a sunny day in Arizona could power NYC through a night time blizzard.

  16. > the article claims the book is 91.91% accurate. That seems like an awfully specific number

    Ah youngsters.

    Back in the old days, businesses had to pay to get included in the book and IF you assume the layout personnel screwed up residential names as often as business names (and ads) then you simply look at the dollars and cents of refunds for screwing up.

    Somehow 8% wrong seems vastly too high, both as in the layout personnel would get fired and the refund department would be a tenth the size of the sales dept. Also I used phone books often enough last century that I recall they failed far less than 8% of the time.

    My gut level guess is the actual written success rate was 99.91% not 91.91% and journalist types unthinkingly copied over and over. A tenth of a percent failure rate sounds like the business acct sales dept data entry personnel were reasonably competent but made a human level of mistakes that the refund department would need to process.

    1. But the article was from 1954, so they had time to improve.

      Surely some inaccuracy was not error, but simply time. Between compiling and release people moved or changed phone numbers for other reasons, and that happened too late to be included. Even people who moved after the book’s release would make the phonebook “inaccurate”.

  17. Project: Stabilize human population (stretch goal: at a sustainable level). i.e. ZPG/NPG
    (Most humane approach for a start would be vastly reducing unintended pregnancies; providing full range of family planning services to all people).

    Bonus – would go a long way toward the world hunger things (or adequate healthy nutrition) others have mentioned. (You can not produce your way out of insufficient nutrition if you are trying to meet an exponentially increasing demand.)

  18. As far as accuracy on clerical processes manually compiled is concerned, we used to expect an initial error rate of 2% on work done in high volume by competent staff.

    That’s why there was always a checking procedure, which with a 2% rate reduced the overall rate to nearly zero.

    On top of that there would normally be a sampling test.

    One thing I remember from those days was how some people had an instinct for something wrong. They could glance at a column of figures and spot discrepancies.

    Possibly a lost skill these days except maybe among programmers.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.