History--computer based information operator terminal system [telecom]

Sometime in my career at Bell Labs (I think in the 1970's, but not sure) our internal newspaper announced a new completely automatic directory-assistance system they were experimenting with. They had it working with the employee phone directory. To use it, you used the telephone keypad to spell out the last name of the person wanted. I don't know if this ever got into production.

That feature is commonly found on small to mid-size PBXes w/ integrated voice mail. Usually you just put in the first 3 "letters" of the person's name, and it iterates the matches, playing the 'personal name' from the voice-mail recording, and asking if this is the person you want.

Of course, when someone forgets to update the line name 'directory', 'funny' things happen.

Some voicemail systems today allow callers to key in the spelling of the desired person's name via the keypad.

Somewhere along way they added the Q and Z to the dial pad, which weren't part of the original lettered dial and QZ weren't used in normal exchange names. For a while some dials had Z over the Operator, why I don't know.

I always wondered about overseas dials. I heard one of the reasons they went ANC was to allow direct dialing overseas, because the letter/ number matchup on overseas dials differed from the US and varied by country. With the widespread use of texting, how is that handled today? Did foreign dials convert to the US format?

Foreign diali9ng came much later than ANC. ANC was promoted as a way to get more office codes as the shortage was loomiing. Wes Leatherock snipped-for-privacy@aol.com snipped-for-privacy@yahoo.com

I believe the original reason for ANC was because the spelling of certain local exchange names wasn't the same as it sounded. For example, most people upon hearing "BAring" would think it's BEaring, not BA, or LOmbard was LUmbard, not LO*. When calls were placed by the operator who plugged into jack it didn't matter, but when people were dialing it did matter.

Also was the problem of people confusing the 0 and O and I and 1.

ANC was pushed fairly early on, in the late 1950s, I think, well before office codes became a shortage in most places. I think New York State might have been an exception to office code shortages, and there they used odd letter codes first as an experiment. I think New Jersey went all ANC as early as 1962.

International Direct Distance Dialing was another reason--I recall seeing pamphlets showing dials of different countries.

*LOmbard was given a new name. However, BAring renames in service to this day as 222. Both in Philadelphia.

Bala-Cynwood was often cited as a prime example. Also note 1 and 0 could not be used in 2L-5N office codes because there were no letters associated with them on the dial. This made unusable a large number of office codes. Wes Leatherock snipped-for-privacy@aol.com snipped-for-privacy@yahoo.com

***** Moderator's Note *****

Wes, I think it's the other way around: letters weren't assigned to "1" or "0" because those numbers were reserved for use in Area Codes (second digit) and Terminating Toll Center codes (First digit), and so they were never considered for use in the "AA" part of "AAN" Exchange codes.

What does the BSTJ say? ;-)

Bill Horne Moderator

It's spelled Bala Cynwyd. Before they moved downtown, Comcast's corporate office was in the GSB Building at One Belmont Avenue, Bala Cynwyd PA 19004.

Well, I think the reason is more fundamental than either of those theories. Initial "1" wasn't used because of the possibility of dialing a false "1" when removing the receiver from the switchhook, and initial "0" was reserved for Operator. There were "no letters associated with them on the dial" precisely *because* they couldn't be used as initial digits.

These restrictions existed as early as 1933, long before area codes were introduced. It was indeed a happy coincidence that "1" and "0" were available when area codes were introduced, but I don't think area codes were even contemplated back in '33.

Here's what Miller had to say in 1933:

| The limitation of eight offices exists because the | number 0 /as a preliminary digit/ is reserved for | reaching the operator, and the digit 1 cannot be | distinguished from an accidental preliminary depression | of the switchhook. Since neither of these can be used to | designate central offices, the eight digits two to nine, | inclusive, remain for this purpose.

Slashes indicate italics in the original.

Source: Kempster B. Miller, M.E. Telephone Theory and Practice: Automatic Switching and Auxiliary Equipment. First Edition. New York: McGraw Hill, 1933, 20.

Neal McLain

The old rule was: Area codes had a 1 or 0 as the second digit, and office codes did not. This was particularly important in places like New Jersey in 1970, where you did not have to dial 1 before a long distance call. The switch looked at the second digit to see how many digits to expect.

Dick

Bells Labs made a study, I believe in the 1950s or so, to see how many false "1s" actually occurred at the start of a csll. After amassing several million examples of dialing, they found not a single case of the false "1". Wes Leatherock snipped-for-privacy@aol.com snipped-for-privacy@yahoo.com

According to the Bell Labs history, the false '1' was on candle stick phones. With the demise of that type of unit in service it was no longer a problem, though some candle sticks remained in service in the

1950s.

When the 202 handset phone came out it was a premium option. Would anyone know when the stopped charging extra for handset units?

I know they gave extra life to candle stick units by retrofiting 302 "F" transmitters and receivers to them, which gave them much better sound quality than the original elements.

The Bell Labs history (published in 1975) says they already forsaw the need to do away with that old rule and were converting switches. But it obviously wasn't a rush thing, I think NJ didn't require a 1 prefix until the mid 1980s. (Today NJ has plenty of area code splits but still only 7 digit dialing for within the area code).

I've heard that before. Nevertheless, that's the reason Miller cited for the no-initial-1 prohibition. Some switches in non-toll-alerting states (Illinois comes to mind) were programmed to simply ignore (absorb) an initial 1. Dialing a 1 didn't even break dialtone.

Folks over on the TCI and Strowger lists have noted that false "1s" were a particular problem with candlestick phones. I guess that makes sense, since it's sometimes difficult to get the receiver out of a candlestick switchhook without bouncing it a time or two. But I suspect that by "the 1950s or so" there weren't any candlesticks in service anyway.

Neal McLain

When area codes were first introduced circa 1952, that's the way it was supposed to be across the entire NANP (which, at the time, consisted of just United States and Canada). W.H. Nunn, writing in September 1952 BSTJ, states that approximately 90 area codes were assigned at the outset. He cites the following rules:

-- Codes with second digit 0 were assigned to single-code states; e.g. NJ (201), FL (305), UT (801).

-- Codes with second digit 1 were assigned to states with two or more area codes; e.g. NY (six codes); CA, PA, IL, TX (four codes each), WI (two codes, 414 and 715).

-- With one exception (maritime provinces), area codes did not overlap any state or provincial boundary.

-- Initial 0 was reserved for Operator, and initial 1 was supposed to be absorbed. [1]

The dial-1-first was a "temporary" artifact of SxS (and possibly other) switching equipment that wasn't capable of examining a second digit to route a call. Dialing the 1 switched the call to some other more-capable switch. Thus, thousands of communities across the US and Canada got used to the idea of dial-1-first. In due time, the initial 1 took on cultural significance. Unfortunately, the cultural significance wasn't consistent from state to state.

NON-TOLL-ALERTING STATES: In come states (notably New York, parts of New Jersey, IBT territory in the Chicago area, and much of California), initial 1 meant "the next three digits you dial will be interpreted as an area code." In these states:

-- Intra-NPA calls, local or toll, were dialed 7D.

-- Inter-NPA calls, local or toll, were dialed 1+10D.

As John Levine wrote in 1993:

TOLL-ALERTING STATES: In other states, initial 1 meant "you are about to dial a toll call." In these states:

-- All intra-NPA local calls were "officially" dialed 7D.

-- Most intra-NPA toll calls were dialed 1+7D.

-- Inter-NPA local calls were dialed 7D (or sometimes 1+10).

-- All inter-NPA toll calls were dialed 1+10D.

There were numerous variations in local dialing plans (both intra- and inter-NPA) due to local numbering conflicts. Even though local calls were "officially" dialed 7D (and were printed as such in directories), many (most?) SxS switches actually required only 4, 5, or 6 digits for local calls. Prepended digits added to produce a 7D number were absorbed at the first selector.

The original plan described by Nunn never happened. By the time all of the old SxS switches had been replaced with newer equipment capable accommodating intermixed central office and NPA codes (without dial-1-first), some N1X and N0X central office codes were already in service in non-toll alerting states. [3] Furthermore, the cultural significance of dial-1-first (whatever it meant) had become so ingrained in our minds that changing it would have been disruptive.

Artifacts of these cultural differences exist to this day. Which, I suppose, is why *all* calls in overlay areas in New York, Illinois, and California require 11D dialing. But overlay areas in toll-alerting states have 10D local and 11D toll.

[1] W. H. Nunn. "Nationwide Numbering Plan." BSTJ, September 1952, 856. [2] John Levine. "Re: 11-Digit Dialing Comes to NYC." TELECOM Digest V22 #251, 23 Jan 2003. [3] On a trip to Los Angeles in 1977, I was dumbfounded to discover 213-305 and 213-314 numbers listed in the white pages.

Neal McLain

Are you sure? NJ has three overlays, and FCC rules require 10D (or

1+10D) dialing for all calls within an overlay.

BTW, which is it, 10D or 1+1D?

Neal McLain

It was up to the individual company and their regulatory agency. Multi-state companies it may have been at different times in each state. Things like this showed the Bell System was not entirely the monolith it was often thought to be. Wes Leatherock snipped-for-privacy@aol.com snipped-for-privacy@yahoo.com

It's 7D in 609, 908 and 856 which are not overlaid.

In the overlaid areas it's supposed to be 10D within the same area (regardless of toll) and 1+10D elsewhere.

R's, John

In 1972, 1+10D dialing was first required in the 213 area (Los Angeles) for calls outside the area. Shortly afterward, n0x and n1x office codes were introduced. Since then, many area code splits have occurred, and in recent years overlays have replaced splits. 1+10D dialing in the home area is permissible throughout California, but required only in overlay areas.

Is that why so many rural places had you dial 1 first for all calls outside a few local exchanges? This, and not the "1 means an area code" method, was standard in most of California until about 1977, when the state PUC mandated "1 means an area code" statewide.

Maybe in metro Los Angeles, "1 meant an area code" much earlier than that, but not in the Bay Area, which allowed 7-digit dialing between its portion of 408 and all of 415 (but required "1" + 7 digits to reach the rest of 408, roughly the part that is now the 831 area code plus Morgan Hill and Gilroy).

After 1977, we went to "1 does not break dial tone" until preparation for NXX area codes began in the '80s.

Not to mention GTE areas, where common services like directory assistance, repair, and time were 11X numbers rather than N11. I don't know if any exchanges like that still exist.

That was true in many Bell areas, too. How about Dallas, Houston, Austin, San Antonio, Fort Worth, Oklahoma City, Tulsa, Wichita plus all the less urban areas in Texas, Arkansas, Oklahoma, Kansas and Missouri? Wes Leatherock snipped-for-privacy@aol.com snipped-for-privacy@yahoo.com