5- vs. 7-digit Dialing (was Area Code Stats)
nmc...@annsgarden.com
> Then there is the case of Centerville, Iowa, population around
> 6,000 for about the last forty years.
> When I was living there, there were phone numbers of the following
> form:
> 856-2xxx, 856-3xxx, 856-6xxx, 856-8xxx, 856-9xxx.
> You never saw phone numbers of the form 856-4xxx or 856-5xxx.
> In town, dialing the last five digits was sufficient. To dial
> neighboring exchanges, seven-digit dialing was required. Three of
> those exchanges were 647-xxxx, 649-xxxx, and 658-xxxx.
> After I moved away, sometime in the 1970s, Centerville opened up the
> 437-xxxx prefix. I always had wondered why a town of that size needed
> two exchanges so early on. But recently, I think I figured it out.
> Centerville may have needed more phone numbers, evidently, but if
> five-digit dialing were preserved, 856-4xxx and 856-5xxx could be
> confused with the nearby exchanges. 6-4xxx could be a Plano or Mystic
> number that was incomplete, or it could be a Centerville number that
> was complete.
> So, I think, 437-xxxx was opened up to provide more numbers and,
> presumably, to preserve five-digit dialing without causing collisions
> with the 6-4xxx or 6-5xxx series. I don't know if Centerville still
> had five-digit dialing by that time. But if it did, 437-nxxx would
> work as long as n != 2 (because there was a neighboring 724- exchange)
> and five-digit dialing would still work.
> It doesn't answer the question of why 6-0xxx, 6-1xxx, or 6-7xxx
> wouldn't have worked, but I am assuming the mechanical switches for
> that exchange may have had limitations. For example, long-distance
> dialing access would be "11", "12" or even "13" depending on some
> arbitrary factor unbeknowst to subscribers of the Iowa State
> Telephone Company (later Continental).
I'll propose a hypothesis.
The Centerville situation you describe is remarkably similar to the
hypothetical step-by-step (Strowger) office described by AT&T in "Notes
on Distance Dialing" [1]. This publication describes "digit absorbing
selectors" -- first selectors having specified levels designated A or
AR:
A = The selector absorbs the specified digit once only; on
the next digit, it "trunks on all levels." This digit
must be dialed once (and only once) in order to reach
certain specified second digits. However, it is absorbed
(ignored) for any other second digit.
AR = The selector absorbs the specified digit repeatedly
unless a digit has been absorbed previously on a level
designated "A".
If my hypothesis is correct, Centerville had a step-by-stop office in
which the first selector was configured as follows:
"5" was type AR, meaning that you could dial 5 repeatedly before
dialing anything else (e.g., 555556-2xxx would reach 856-2xxx;
5555555555555550 would reach the operator).
"6" was type A, meaning that you had to dial 6 once (with or
without any previous 5s or 8s) in order to reach the 6-2xxx,
6-3xxx, 6-6xxx, 6-8xxx, and 6-9xxx local numbers or 658-xxxx
numbers in a neighboring community. But 6-4... would reach
numbers beginning with 4; 6-0 would reach the operator; etc.
"8" was type AR, meaning that you could dial 8 repeatedly before
dialing anything else (e.g., 8888886-3xxx would reach 856-3xxx).
Or, for that matter, 5858585858586-2xxx would reach 856-2xxx, or
5555588885858858885888888586-0 would reach the operator.
Therefore, the dialing plan would have been as follows. In this list:
[N] is a footnote reference.
<85> are the two "AR" digits (although any combination
of 5s and 8s could be substituted, or they could be
omitted altogether).
(6) is the "A" digit (required if the second digit is
2, 3, 5, 6, 8, or 9; absorbed and ignored if the second
digit is 1, 4, 7, or 0).
<85>(6)11... Long distance
<85>(6)12... Long distance
<85>(6)13... Long distance
<85>(6)1N ??? [2]
<85>(6)2xxx Local numbers in form 856-2xxx
<85>(6)3xxx Local numbers in form 856-3xxx
<85>(6)437-xxxx Local numbers in form 437-xxxx [3]
<85>(6)47-xxxx Nearby numbers in the form 647-xxxx [4]
<85>(6)49-xxxx Nearby numbers in the form 649-xxxx [4]
<85>(6)58-xxxx Nearby numbers in the form 658-xxxx [5]
<85>(6)6xxx Local numbers in form 856-6xxx
<85>(6)7... Blank level (or possibly 724?)
<85>(6)8xxx Local numbers in form 856-8xxx
<85>(6)9xxx Local numbers in form 856-9xxx
<85>(6)0 Operator
All of this is, of course, just hypothesis, so further comments are
welcome!
Similar situations existed in small and mid-sized communities throughout
the Bell System during the 50s, 60s, and 70s [6].
A final question would be: how is a first selector configured for A and
AR levels? I don't know the answer to this one; perhaps Roger Conklin
or Joe Stevens can explain it to us.
Footnotes:
[1] Appendix A, Section 4, "Typical Trunking Diagrams for Step-by-Step
Offices." "Notes on Distance Dialing." AT&T Engineering and Network
Services Department, Systems Planning Section, 1975.
[2] Possibly other codes such as information, revertive calling, etc?
[3] The choice of this particular NXX code (437) would have allowed the
introduction of new numbers in Centerville before the 856 step-by-step
equipment was replaced with newer equipment (crossbar or ESS). However,
I would surmise that the 437 equipment itself was either crossbar or ESS
(to the best of my knowledge, by "sometime in the 1970s" Bell was no
longer installing any new step-by-step equipment). So seven-digit
dialing would have been required for any local call from a 437 number
even if 856-to-856 calls remained 5-digit.
[4] Which means that you could have reached 647-xxxx or 649-xxxx numbers
from Centerville by dialing only the last six digits. Did you ever try
this?
[5] In this case, the full seven-digit number (including the 6) would
have been required. Omitting the 6 would cause 5 (the AR level) to be
the first digit, so it would have been absorbed and ignored.
[6] One example: Ann Arbor, Michigan. In Ann Arbor, 6 was type AR and 8
was type A. My previous post about this situation is in the archives at
<http://groups.google.com/groups?q=2-4401&hl=en&lr=&ie=UTF-8&selm=200012090906_MC2-BDFD-FCCD%40compuserve.com&rnum=1>.
Neal McLain
nmc...@annsgarden.com
Mike O
>> After I moved away, sometime in the 1970s, Centerville ...
I vaguely remember a "rule" the phone company had when they made area
codes available to the public dial system (1960?) . The primary rule
was that the exchange (prefix) could not have a zero or one, that was
reserved for area codes. But, they also said that future area codes
(when they ran out of zero and one, in a projected 25 years) would use
2 and 3, as the center digit. Then 4 and 5, etc. I don't know how
valid this was, or if it was just an ":urban legend" from that era,
but it may shed some light on numbering schemes.
Cheers,
Mike
[TELECOM Digest Editor's Note: Area codes began being used in the
1950's in a few places, even while many places were still entirely
manual in operation. The rule was no zeros or ones as the middle digit
in prefixes, and no zeros as the right, or last digit in a prefix.
That's largely because the early or original area codes all had zeros
or ones as the middle digit. There was also a rule that states which
had only one area code for the entire state had a number which used
zero as the middle digit; a second or subsequent area code in the
same state had one as the middle digit. PAT]
Mark Roberts
> Mark Roberts <mark...@hotmail.com: wrote:
>> Then there is the case of Centerville, Iowa, population around
>> 6,000 for about the last forty years.
>> When I was living there, there were phone numbers of the following
>> form:
>> 856-2xxx, 856-3xxx, 856-6xxx, 856-8xxx, 856-9xxx.
>> You never saw phone numbers of the form 856-4xxx or 856-5xxx.
[...]
>> So, I think, 437-xxxx was opened up to provide more numbers and,
>> presumably, to preserve five-digit dialing without causing collisions
>> with the 6-4xxx or 6-5xxx series.
[...]
>> It doesn't answer the question of why 6-0xxx, 6-1xxx, or 6-7xxx
I forgot to mention that there were special service codes beginning
with "7": "71" for directory assistance, "70" for repair service,
"75" for the mobile operator.
> I'll propose a hypothesis.
Thanks, very interesting. There is one minor fly in the ointment in
that this wasn't a Bell area (it was Continental by then). But I
would think the 437 exchange would have had newer equipment.
> [3] The choice of this particular NXX code (437) would have allowed the
> introduction of new numbers in Centerville before the 856 step-by-step
> equipment was replaced with newer equipment (crossbar or ESS). However,
> I would surmise that the 437 equipment itself was either crossbar or ESS
> (to the best of my knowledge, by "sometime in the 1970s" Bell was no
> longer installing any new step-by-step equipment). So seven-digit
> dialing would have been required for any local call from a 437 number
> even if 856-to-856 calls remained 5-digit.
This reminds me of the situation that prevailed in my GTE hometown in
Missouri, where the original three exchanges of GIbson 2, 3, and 9
were surrounded by newer exchanges starting in 1971. Until the whole
mess was swept away in 1986, 442, 443, and 449 customers could reach
each other with the last five digits only. The other exchanges were
started with "4", "6", or "8". Customers in the newer 445 exchange had
to dial all seven digits for all local calls. The peripheral exchanges
sometimes had trouble dialing each other. Even into the early 1980s, I
remember having to resort to having an operator complete calls between
445 and 474 numbers.
There was also an unusual number for time and temperature: 655. For
customers in the "6" exchanges, that number was instead 446. Both
were replaced in 1986 by 449-0655.
Mark Roberts | "In 2001, Google bought the Usenet discussion forums."
Oakland, Cal.| -- Associated Press dispatch published in the
NO HTML MAIL | San Francisco Chronicle, 2-19-2003