Pages: (1)_(2)_(3)_(4)_(5)_(6)_(7)_(8)_(9)_(10) (Communicating) (Soundwaves) (Life at Western Electric) next page -->

<--back to telephone history series

<-- back to mobile telephone history series 

Digital Roots, a discussion

David Robertson:

"As part of the work for my Paper for the IEE, I'm trying to get to grips with the 'antecedents' of PCM - the ideas and technologies on which Reeves may have drawn. We know that sampling had been suggested by Miner and Squier, while Morse or his assistant Vail came up with what we now think of as regeneration."

"What's much less clear -- at least to me -- is whether anyone else had had the idea of transmitting speech in a 'telegraph-like' way - perhaps the key feature of PCM. Some writers (including, interestingly, Reeves' boss Maurice Deloraine in 'When Telecoms and ITT were Young') have suggested that Page, Bourseul and Reis were attempting to do exactly that when, in their early experiments with what eventually became telephony, they sought ways to turn speech into a form suitable for transmission over the telegraph network."

"It appears to me that such attempts were doomed to failure as we know they largely were are 'digital' only in the very limited, technical sense that the telegraph was all they knew about and hence was the model they were exploring. Reeves, by contrast, came after Bell and hence was proposing to sample the by-now-existing 'analogue' signal and turning these samples into 'digital', that is, a telegraph form. This appears to me a far more subtle and more complex task. In making this suggestion, I don't want to appear to denigrate what Page et al did, but I really don't see them as part of the 'digital' story."

"What do you think ?"

Tom Farley:

"Yes, you are totally correct. The early inventors, those before Bell, were all on the wrong path, thinking they could reproduce speech in those early years by making and breaking a circuit, just like with the telegraph. They were a million miles from success, just plodding along, electricians in a field that needed people familiar with electricity and acoustics. They weren't on this path deliberately but blindly, so I would pay them little attention in a history of digital beginnings. Let me give one example, this taken from my own writing in my telephone history series:"

In 1861 Johann Phillip Reis completed the first non-working telephone. Tantalizingly close to reproducing speech, Reis's instrument conveyed certain sounds, poorly, but no more than that. A German physicist and school teacher, Reis's ingenuity was unquestioned. His transmitter and receiver used a cork, a knitting needle, a sausage skin, and a piece of platinum to transmit bits of music and certain other sounds. But intelligible speech could not be reproduced. The problem was simple, minute, and at the same time monumental. His telephone relied on its transmitter's diaphragm making and breaking contact with the electrical circuit, just as Bourseul suggested, and just as the telegraph worked. This approach, however, was completely wrong.

Reproducing speech practically relies on the transmitter making continuous contact with the electrical circuit. A transmitter varies the electrical current depending on how much acoustic pressure it gets. Turning the current off and on like a telegraph cannot begin to duplicate speech since speech, once flowing, is a fluctuating wave of continuous character; it is not a collection of off and on again pulses. The Reis instrument, in fact, worked only when sounds were so soft that the contact connecting the transmitter to the circuit remained unbroken. Speech may have traveled first over a Reis telephone, however, it would have done so accidentally and against every principle he thought would make it work. And although accidental discovery is the stuff of invention, Reis did not realize his mistake, did not understand the principle behind voice transmission, did not develop his instrument further, nor did he ever claim to have invented the telephone.

Don Kimberlin:

"I think the people working on digital owe somethng to the early pioneers. I'd say due to the close telegraphic relations at the beginning of telephony, those who were developing the telephone network in the early Bell 'labs', then simply the 'engineering department of AT&T, were seeking a method not necessarily to send speech over telegraph lines (although that is a tantalizing notion as to their purpose), but like Morse, they wanted a regenerative way to extend transmission lengths. That would be a digital method. Remember, this was in the time period when DeForest's Audion was relatively undeveloped when around 1914 Bell got analog audio amps developed enough, by understanding and employing negative feedback to stablize them, to be able to make a transcontinental connection. When you match that with the date of the earliest Bell papers talking about how to sample and encode speech for digital transmission, you can see where analog vs. digital was simply the Bell interests playing all their bets as to how to get a national long distance network functioning. That's why I lean toward giving Bell credit for the deep origins of digital telephony."

"As to Reeves, my American-biased mind gives heroic credit to Alec Reeves for constucting the first PCM channel bank at Standard Telephone Laboratories, but not for inventing the processes it operated with. That has a root in mathematical papers dating to 1914 at the Engineering Labs that I alluded to above. As I wrote, even from the outset, the Bell people wanted to be able to have a regenerative telephone network, and always looked upon analog transmission as a necessary evil. The main thing that delayed their use of digital transmission as a fielded technology until 1964 was economics. And, of course, it was the serendipitous stumbling upon the transistor (while looking for lower cost solid state rectifiers, not improved transmission), that made digital transmission finally become a reality. Reeves' digital channel bank, while a marvelous accomplishment, was an overly expensive demonstration project."

"If one wants British accomplishment to point to, make a visit to Bletchley Park and learn about how British innovation created the world's first programmable digital computer, Colossus, two years before the Yanks built ENIAC. It was purely and simply MI6 secretiveness that's kept the world from knowing about it. Now, the Official Secrets Act blanket has been raised from Colossus. MI6 thought they'd burned all the records on it, but they neglected to remember the Yanks had been given info on Colossus. Today, a group of veterans have rebuilt one, WWII electronic valves and all. (They figure its computing horsepower roughly compares to one of today's Pentium PC's.) See http://www.bletchleyPark.org (external link)."

David Robertson

"I see that we nearly worked in the same company, in that you were with ITT from 1968 - 1973 while I joined STC, its main UK subsidiary, in 1975. (I was once in Geneen's office, but never met him). This means, inter alia, that you overlapped with Reeves while I just missed him."

"Briefly, before I accept the notion that the world and Ken Cattermole are wrong and that Alec Reeves didn't invent PCM, I will need a whole lot more proof than is provided here. I am, of course, aware of some of the theoretical papers by Goodall, Pierce, Shannon, and others, but am not yet persuaded these constitute an 'invention' , any more than I have yet been able to agree with those who say Paul Rainey's electromechanical TV facsimile patent of 1926 'is' PCM either."

"As to Bletchley: for reasons with which I will bore you tomorrow, it is not a place I ever intend to visit again. There's a person there who has done immense damage to important work there, not least the attempt to restore Colossus."

"Whether Colossus was a computer sounds like another story for another day, too. My money, for what it's worth, is on Maurice Wilkes at Cambridge!"

"To be truly radical before I sign off, I believe Reeves was a hugely more important figure than Turing, but there are people here who will kill rather than accept such a view."

Tom Farley

"Your e-mails points to the hard to establish date when the electrical age in telephony ended and the electronics era began. That time swirls around when some kind of electron tube became employed, when we went from manipulating electricity in the main to managing a specific property of it, the electron. Or can we date it to The Edison Effect when that was noticed but not investigated at the time? Without getting out my books for dates, I recall that AT&T's folks were proceeding slowly until they inspected what DeForest had stumbled upon. DeForest had by trial and error experimenting, truly the method of rudimentary electricians, backed into a great invention. I remain convinced he never truly understood how it worked, even after people like Armstrong explained it to him. But I ramble and am now far away from the topic. "

Don Kimberlin

"Tom, that date of when telephony first used electronics is, in my mind, rather firmly fixed to the date that AT&T first added amplification to create the first transcontinental phone circuit. I, too, would have to get the books out, but recall that was around 1914 or 1915. And, here's a little quirk of analog/digital and telephony/telegraphy relations: The next major use of electronics by Bell was in their very first 'carrier' system. You may have noted that over the years, they assigned successive letter "types" to each new carrier telephone development."

"Most people never look deeply enough to find out that Type A Carrier was actually a system that produced several telegraph channels placed above the 3 kHz limit of a voice channel on wire. It wasn't until later development that they placed a second speech channel above baseband speech."

David Robertson:

Don and Tom,

"Many thanks for the messages and its various successors. I have set the cat among the pigeons!"

"It seems to me we need to distinguish carefully between the separate questions of 1) when switching and transmission became electronic, as distinct from electromechanical, and 2) when they became digital, as opposed to analogue. And there's another question to explore."

"Reeves' champions, like Ken Cattermole, believe only he invented full digital working or PCM, because he brought together sampling, quantizing, transmission in binary, regeneration and multiplexing. Certainly telegraphs are digital but only in the most limited technical sense. They turn current on and off to send a message, nothing more. How much does Reeves or anyone involved in PCM owe early telegraphy as an inspiration?"

"I cannot remember whether I referred you to an interesting paper by Norwegian academic Lars Lundheim. I cannot find the actual address but it can be found by searching on 'LarsTelektronikk02.pdf'. As I understand it, he argues that the final completion of much of the conceptual work of Shannon and Nyquist had to await the practical, engineering work of Reeves."

"I would be very interested in having details of the specific Bell papers to which you've referred. My reading of Nyquist's 1928 paper on telegraphy is that it cannot be taken as explicitly prefiguring PCM - indeed it doesn't mention the word 'speech' once - though some of the theoretical antecedents are there. What other documents are you referring to? Where can they be found?"

"A friend of Reeves' with whom I've been in touch says he would have been amused by all this interest in who-did-what. He wasn't in the least interested in precedence or fame but just getting on with the job. There is a sense in which only biographers, patent lawyers and (sometimes) politicians worry about how to ascribe inventions to people. But, as Tom has pointed out to me many times in respect of Bell and the telephone, 'invention' means getting the thing into a meaningful and useful state. This is why Tom won't 'buy' the claims of Reis or Bourseul. Using that parallel, there doesn't seem to me any doubt that Reeves 'invented' PCM and thereby the digital transmission of speech (and later music and pictures). I would go further and suggest his claim is more unique than, say, Marconi's (with respect to people like Lodge) and maybe even Bell. But that, I realise, is fighting talk!

Kind regards,

David

For more on digitizing and PCM, click here. (internal link)

/back to telephone history series/ /back to mobile telephone history series