Bell Labs effectively became outsourced R&D for many technologies post WW2 like solar PV as you pointed out.
I also think one challenge with recreating it today is the financial context. Most of Bell's inventions came prior to the financialization of corporate America. Before MBAs ran companies to quarterly results, before buyouts ran companies into the ground, before buybacks sapped $10s of billions of potential R&D from previously innovative companies.
The only types of companies today which could hope to do exploratory R&D on any scale approaching Bell would have to be founder-run with full board support and preferably privately held. Elon's companies, Zuck, Bezos and James Dyson's R&D efforts come to mind.
In WWII he was a Navy Combat Engineer aka a Seabee aka a construction foreman for building concrete bunkers under fire.
After the war he joined the Navy reserve and moved to California. At one point, he was called up to help pour concrete for a school.
Then came Korea, and he received mobilization orders. Report to Seoul for duty as a JaG.
Well, that didn’t sound right, as he’d never had any legal training whatsoever, but if the Navy tells you to show up, you show up. So he showed up in Seoul ready for duty. And pointed out the error to his CO.
And his CO said that no, that’s not right. It says right here in your service record that you were “instrumental in building the foundation for a new school of law”
So, there had been a screwup. But, the Navy had their man that they asked for. And if he wasn’t ready for the job, well, that’s what Navy training is for!!!
So they taught him how the navy procurement system works. What documents you need. What clearances must be in place. And he spent the war in Seoul supporting Navy contracting.
After the war, he went straight to Bell Labs. And told them “hey! I’m a navy procurement officer. So, I’m an expert in what you need to do if you want to sign contracts with the US Navy. What requirements you have to meet, what clearances must be obtained and how to obtain them. And I know the people involved and how to work with them. That kind of stuff. Do you guys have any interest in someone like that?”
So, he ended up setting aside the concrete pouring and worked for Bell Labs for the rest of his career.
When it comes to engineers, all you really have to do is to offer them the chance to do the very best work of their lives. You don't have to pay them that much, and you don't have to give them a fancy title -- Bell Labs famously had none -- just the opportunity. This is true of all the support engineers as well as the reseachers. The question then becomes 'what sort of organisation do we build so that engineers get that opportunity'?
If any of you get a chance to build such a thing, can I ask you to please read Charles Perrow's *Complex Organizations* (Third Edition)? It's old. Perrow is dead now, alas. But it is a great survey work on the structure of bureaucracies. As discussed in the chapter on 'the Neo Weberian Model', for a research organisation, you want to be building something that can handle complex (as opposed to linear) interactions with loose (as opposed to tight) coupling. Unfortunately, the base sort of bureaucratic model that is well understood is that supporting the assembly line, which is an example of something that has linear (as opposed to complex) interactions with very tight (as opposed to lose) coupling. If you screw up your bureaucratic design, you will fail, no matter how many brilliant engineers you recruit.
p.s. -- Perrow is supremely readable, too. This book is on my short list of 'things I wish that everybody had read', but they won't, of course. But it is directly revelant to this proposed project, so maybe I can get some takers here?
No, I don't think it's true these days that you "don't have to pay them much." Most engineers and scientists you could plausibly work at a neo-Bell Labs could just as well go make a quarter or half a million dollars at a FAANG company or in AI, and that's a real, serious incentive. In fact I've spoken specifically to people at Sandia, one of the most closely comparable things we have left, who lament that it's impossible to do their work when every new hire leaves for Google three years out of the Ph.D.
Well, the value of legacies in engineering has fallen quite far compared to the days of Bell Labs and even Perrow, so the FAANG dollars and what that gets you is a powerful motivator indeed.
I really should have this written down somewhere, instead of on a metaphorical list. But David Pye's *The Nature and Art of Workmanship* and *The Nature and Aesthetics of Design* (2 volume set can be read in either order, but do read both) makes the list, as does Nassim Taleb's *Fooled by Randomness* and George Polya's *Mathematical Discovery* and *Teaching Learning and Learning Teaching*. Ernst Mayr's *This is Biology*. Eugene Ferguson's *Engineering and the Mind's Eye*.
What's on your list? (this is a game everybody can play! :) )
That's a cool list. I honestly don't recall reading any non-fiction except for a few self help books. I've mostly read fiction and textbooks. As far as textbooks, I feel that the Course of Theoretical Physics by Landau should be essential for a lot of physical scientists. There may be a few articles I read pertaining to the philosophy of Physics or Math that I found enlightening, which I can scrounge up if you're interested.
I think you're missing the key benefit of being a regulated monopoly: every dollar spent on R&D was automatically guaranteed to be profitable, because they were allowed to simply charge customers more. For every dollar spent on R&D, regulators allowed AT&T to charge customers, say, $1.10.
Bell Labs was funded through a 1% tax on Bell's overall revenue, as well as a claim on a portion of Western Electric's revenue, so yes, AT&T was able to recoup Bell Labs' expenses. It's not clear to me that they were able to rate base those expenses (which would enable them to earn a profit on those expenses), or if they were counted as operating expenses (which would simply mean cost recovery).
Claude.ai tells me they were able to rate base expenses for Bell Labs (not that that definitifely answers the question). Either way, recovering the cost and also getting the downstream benefits (like patent licensing) is a pretty cushy deal.
In late 1969, our Denver “underground” radio station obtained an audio tape produced by Bell Lab engineers called “The Endlessly Descending Glessenda.” It was quite spacey.
I might also say that my original physics direction was low-temperature solid state physics and Bell Labs would have been an ideal direction. While finishing my PhD in solid state physics I began working in high-energy particle astrophysics as well. On graduation, I followed that path rather than Solid State. I can’t say I never looked back but the fullness of time has confirmed my decision.
Yes. NASA Goddard Space Flight Center (GSFC) is the only NASA Center the main mission of which is science. Indeed it was the first NASA center to have a significant science focus. I joined GSFC in 1990 after 5 years as a postdoctoral researcher in experimental high-energy particle astrophysics. I found the GSFC staff to be exceptionally dedicated and gifted. I also found the GSFC leadership to be very committed to research excellence which inspired me. For 19 years, my office was on the ground floor of GSFC building 2 which was the original science building at the center. It was often said that anyone who was anyone in high-energy astrophysics, or really in any discipline of astrophysics or astronomy, would eventually pass through the hall by my office. For a young, active researcher, that was and is an exceptional environment in which to work. Building 2 is gone sadly, victim of federal building space regulations, but its spirit continues forward.
Bell Labs could have done even better if they could have hired a more diverse group of scientists and engineers. People from different backgrounds look at problems differently. Bell Labs did have the advantage of being able to attract international talent, but even they were from only a small part of the population.
> Bell Labs was gradually carved apart through constant mergers and spinoffs to different organizations, and it has never achieved the heights it did in the 1950s through the 1980s.
Where does "1950s" come from? The entire following retold history suggests that Bell Labs was "so on top of its game it got the Vertigo" from a much earlier period.
Interesting, in a never meet your heroes way. Given the large number of unfair advantages they had, if anything they seema little unimpressive now. The failure mode of doing r&d that never gets deployed is a common one but once that modern corporations are much better at avoiding, I think.
The MIT Radiation Laboratory was called the Rad Lab, creating some confusion with Ernest Lawrence’s nuclear physics lab at UC Berkeley, which was often known by the same name. In my solid state physics work, I used the “Rad Lab series” of books that set out virtually the whole foundation of microwave circuits and applications. I also worked directly with the UCB Rad Lab so I am well aware of the two.
On a different subject, many parallels can be drawn between Bell Labs and the MIT Radiation Laboratory and later MIT Lincoln Labs. The Rad Lab was formed during WW II to advance microwave radar but had a deep influence on the development of much of the technology that enabled everything from weather radar to radio astronomy.
I should further have said that I am still at GSFC, continuing my research, several years after I could have comfortably retired. That is mostly because I love my research and I am dedicated to the continued health of my research group: GSFC high-energy cosmic radiations (HECR) experimental research. The GSFC HECR group began in the 1970s and has done remarkable work. I have led it since 2007.
I wanted to work at Bell Labs, of course during its most productive years. Instead, I ended up working for NASA at Goddard Space Flight Center. After 34 years I can’t imagine Bell Labs being a better place to work.
Bell Labs effectively became outsourced R&D for many technologies post WW2 like solar PV as you pointed out.
I also think one challenge with recreating it today is the financial context. Most of Bell's inventions came prior to the financialization of corporate America. Before MBAs ran companies to quarterly results, before buyouts ran companies into the ground, before buybacks sapped $10s of billions of potential R&D from previously innovative companies.
The only types of companies today which could hope to do exploratory R&D on any scale approaching Bell would have to be founder-run with full board support and preferably privately held. Elon's companies, Zuck, Bezos and James Dyson's R&D efforts come to mind.
Big 'Image Not Found' after the image that is the list of Bell Labs Nobel Prize winners.
Thanks, fixed.
My grandfather worked at Bell Labs. By accident.
In WWII he was a Navy Combat Engineer aka a Seabee aka a construction foreman for building concrete bunkers under fire.
After the war he joined the Navy reserve and moved to California. At one point, he was called up to help pour concrete for a school.
Then came Korea, and he received mobilization orders. Report to Seoul for duty as a JaG.
Well, that didn’t sound right, as he’d never had any legal training whatsoever, but if the Navy tells you to show up, you show up. So he showed up in Seoul ready for duty. And pointed out the error to his CO.
And his CO said that no, that’s not right. It says right here in your service record that you were “instrumental in building the foundation for a new school of law”
So, there had been a screwup. But, the Navy had their man that they asked for. And if he wasn’t ready for the job, well, that’s what Navy training is for!!!
So they taught him how the navy procurement system works. What documents you need. What clearances must be in place. And he spent the war in Seoul supporting Navy contracting.
After the war, he went straight to Bell Labs. And told them “hey! I’m a navy procurement officer. So, I’m an expert in what you need to do if you want to sign contracts with the US Navy. What requirements you have to meet, what clearances must be obtained and how to obtain them. And I know the people involved and how to work with them. That kind of stuff. Do you guys have any interest in someone like that?”
So, he ended up setting aside the concrete pouring and worked for Bell Labs for the rest of his career.
When it comes to engineers, all you really have to do is to offer them the chance to do the very best work of their lives. You don't have to pay them that much, and you don't have to give them a fancy title -- Bell Labs famously had none -- just the opportunity. This is true of all the support engineers as well as the reseachers. The question then becomes 'what sort of organisation do we build so that engineers get that opportunity'?
If any of you get a chance to build such a thing, can I ask you to please read Charles Perrow's *Complex Organizations* (Third Edition)? It's old. Perrow is dead now, alas. But it is a great survey work on the structure of bureaucracies. As discussed in the chapter on 'the Neo Weberian Model', for a research organisation, you want to be building something that can handle complex (as opposed to linear) interactions with loose (as opposed to tight) coupling. Unfortunately, the base sort of bureaucratic model that is well understood is that supporting the assembly line, which is an example of something that has linear (as opposed to complex) interactions with very tight (as opposed to lose) coupling. If you screw up your bureaucratic design, you will fail, no matter how many brilliant engineers you recruit.
p.s. -- Perrow is supremely readable, too. This book is on my short list of 'things I wish that everybody had read', but they won't, of course. But it is directly revelant to this proposed project, so maybe I can get some takers here?
No, I don't think it's true these days that you "don't have to pay them much." Most engineers and scientists you could plausibly work at a neo-Bell Labs could just as well go make a quarter or half a million dollars at a FAANG company or in AI, and that's a real, serious incentive. In fact I've spoken specifically to people at Sandia, one of the most closely comparable things we have left, who lament that it's impossible to do their work when every new hire leaves for Google three years out of the Ph.D.
Well, the value of legacies in engineering has fallen quite far compared to the days of Bell Labs and even Perrow, so the FAANG dollars and what that gets you is a powerful motivator indeed.
What is the rest of your short list?
I really should have this written down somewhere, instead of on a metaphorical list. But David Pye's *The Nature and Art of Workmanship* and *The Nature and Aesthetics of Design* (2 volume set can be read in either order, but do read both) makes the list, as does Nassim Taleb's *Fooled by Randomness* and George Polya's *Mathematical Discovery* and *Teaching Learning and Learning Teaching*. Ernst Mayr's *This is Biology*. Eugene Ferguson's *Engineering and the Mind's Eye*.
What's on your list? (this is a game everybody can play! :) )
That's a cool list. I honestly don't recall reading any non-fiction except for a few self help books. I've mostly read fiction and textbooks. As far as textbooks, I feel that the Course of Theoretical Physics by Landau should be essential for a lot of physical scientists. There may be a few articles I read pertaining to the philosophy of Physics or Math that I found enlightening, which I can scrounge up if you're interested.
I am always interested in what other people think of as 'essential reading' -- or indeed just 'really interesting stuff'.
I've scoured my records, and would generally recommend the following, though the last four are more for people interested in science:
Man's Search for Meaning - Frankl
The Nightmare that is a Reality - Koestler
A Mathematician's Lament - Lockhart
You and Your Research - Hamming
Cargo Cult Science Speech - Feynman
Birds and Frogs - Dyson
Chaos in the Brickyard - Forscher
The Evolution of the Physicist's Picture of Nature - Dirac (have not finished)
The book by Lockhart is relevant to anyone- educators, parents, students.
Structures: Or Why Things Don't Fall Down - J.E. Gordon
The New Science of Strong Materials - J.E. Gordon
On the Psychology of Military Incompetence - Norman F. Dixon
Our Own Worst Enemy - Norman F. Dixon
Money: Whence it Came, Where it Went - J.K. Galbraith
The Dawn of Everything - D. Graeber & D. Wengrow
I had forgotten about J.E. Gordon! Thank you.
I think you're missing the key benefit of being a regulated monopoly: every dollar spent on R&D was automatically guaranteed to be profitable, because they were allowed to simply charge customers more. For every dollar spent on R&D, regulators allowed AT&T to charge customers, say, $1.10.
Bell Labs was funded through a 1% tax on Bell's overall revenue, as well as a claim on a portion of Western Electric's revenue, so yes, AT&T was able to recoup Bell Labs' expenses. It's not clear to me that they were able to rate base those expenses (which would enable them to earn a profit on those expenses), or if they were counted as operating expenses (which would simply mean cost recovery).
Claude.ai tells me they were able to rate base expenses for Bell Labs (not that that definitifely answers the question). Either way, recovering the cost and also getting the downstream benefits (like patent licensing) is a pretty cushy deal.
In late 1969, our Denver “underground” radio station obtained an audio tape produced by Bell Lab engineers called “The Endlessly Descending Glessenda.” It was quite spacey.
I might also say that my original physics direction was low-temperature solid state physics and Bell Labs would have been an ideal direction. While finishing my PhD in solid state physics I began working in high-energy particle astrophysics as well. On graduation, I followed that path rather than Solid State. I can’t say I never looked back but the fullness of time has confirmed my decision.
Yes. NASA Goddard Space Flight Center (GSFC) is the only NASA Center the main mission of which is science. Indeed it was the first NASA center to have a significant science focus. I joined GSFC in 1990 after 5 years as a postdoctoral researcher in experimental high-energy particle astrophysics. I found the GSFC staff to be exceptionally dedicated and gifted. I also found the GSFC leadership to be very committed to research excellence which inspired me. For 19 years, my office was on the ground floor of GSFC building 2 which was the original science building at the center. It was often said that anyone who was anyone in high-energy astrophysics, or really in any discipline of astrophysics or astronomy, would eventually pass through the hall by my office. For a young, active researcher, that was and is an exceptional environment in which to work. Building 2 is gone sadly, victim of federal building space regulations, but its spirit continues forward.
Would that be with or without the mandatory DEI these days? And with or without all the backdoors for the FedPigs without any subpoenas?
Bell Labs could have done even better if they could have hired a more diverse group of scientists and engineers. People from different backgrounds look at problems differently. Bell Labs did have the advantage of being able to attract international talent, but even they were from only a small part of the population.
IIRC, it was difficult for aliens, even with good American PhDs, to get hired there.
Bullshiite!
That proves my point.
Great article, but I have a nitpick.
> Bell Labs was gradually carved apart through constant mergers and spinoffs to different organizations, and it has never achieved the heights it did in the 1950s through the 1980s.
Where does "1950s" come from? The entire following retold history suggests that Bell Labs was "so on top of its game it got the Vertigo" from a much earlier period.
Interesting, in a never meet your heroes way. Given the large number of unfair advantages they had, if anything they seema little unimpressive now. The failure mode of doing r&d that never gets deployed is a common one but once that modern corporations are much better at avoiding, I think.
The MIT Radiation Laboratory was called the Rad Lab, creating some confusion with Ernest Lawrence’s nuclear physics lab at UC Berkeley, which was often known by the same name. In my solid state physics work, I used the “Rad Lab series” of books that set out virtually the whole foundation of microwave circuits and applications. I also worked directly with the UCB Rad Lab so I am well aware of the two.
On a different subject, many parallels can be drawn between Bell Labs and the MIT Radiation Laboratory and later MIT Lincoln Labs. The Rad Lab was formed during WW II to advance microwave radar but had a deep influence on the development of much of the technology that enabled everything from weather radar to radio astronomy.
I should further have said that I am still at GSFC, continuing my research, several years after I could have comfortably retired. That is mostly because I love my research and I am dedicated to the continued health of my research group: GSFC high-energy cosmic radiations (HECR) experimental research. The GSFC HECR group began in the 1970s and has done remarkable work. I have led it since 2007.
I wanted to work at Bell Labs, of course during its most productive years. Instead, I ended up working for NASA at Goddard Space Flight Center. After 34 years I can’t imagine Bell Labs being a better place to work.
Could you expand on NASA Goddard? Why was it so great as a workplace?
https://open.substack.com/pub/emilyalexandraguglielmo/p/my-first-public-blogpost?r=2mtps5&utm_medium=ios