Beyond commonsense and physical access controls, is the system password the oldest still in active use security feature? By physical access, I mean for example leaving the system within a secure space.


In ant colonies, specialized ants act as guards and prevent other ants from entering some areas unless they "present" the appropriate, colony-specific pheromone. This is the password model, and ants have been doing that since (at least) mid-Cretaceous, 110 million years ago. So I daresay that the "password" is an ancient security system.

In the realm of human endeavors and technology, cars have featured keys (as security features, i.e. controlling whether the car can be operated at all) since the 1920s.

  • 1
    You win, since a signal in my opinion is "information technology" regardless if it's of internal, or external, origin -- the scenario is clearly security oriented, and still in active use; plus, human involvement was not stated as a requirement. Also of interest to me is that if you took an "enemy" ant, and gave them the colony-specific pheromone, they'd be allowed in; this based on the fact that living ants coded with a pheromone only dead ants give off are treated as dead, and taken to the dead ant pile; at which point the "undead" ant cleans off, and rejoins the colony. – blunders Jun 6 '11 at 20:56

If you were to consider electronic or electro-mechanical systems, I doubt it would be the password. Instead it would be ciphers, which were in military use.

If you have to go farther back in history and consider non-electronic systems, there is of course, the concept of the security seal, the first being the bulla, whose modern equivalent is the digital certificate. And there is bookkeeping which might have come in around the same time, which ostensibly did not exist without the bulla.

You'll find all of this discussed in Security Engineering, by Ross Anderson. The book, however, carries far more useful knowledge than trivia.

  • +1 Ciphers are still based on the old principles of substitution (like the caesar cipher) and transposition (like the greek scytale), they have just become more elaborate. – starblue Jun 6 '11 at 14:31
  • which is older? cipher or steganography? – Lie Ryan Jun 6 '11 at 18:29
  • I think ciphers are older than steganography. But I might be wrong. After all, both date far back to BCE. – Vineet Reynolds Jun 6 '11 at 20:32
  • Ciphers are generally speaking based on keys (i.e. passwords). Know the right key/password and you can decipher it. So i guess crypto and passwords go hand in hand. – Henri Jun 7 '11 at 18:09
  • @Henri, you are partially correct. Ciphers are based on keys, but keys are not necessarily the same as a password. Passwords (ignoring ants) have lesser entropy than typical keys, and usually are sent to key-derivation functions, if they have to function as keys. KDFs usually strengthen the key by adding a salt, and "stretch" the key by performing several iterations. – Vineet Reynolds Jun 7 '11 at 18:17

"Security" is traditionally defined to include the CIA triad: confidentiality, integrity and availability.

Long before computer users had to worry about confidentiality, beyond locking the room that held the computer, they had to make a mechanism that provided integrity and availability - the bloody thing had to work.

Getting an astrolabe, Antikythera mechanism (thanks for the reminder, @blunders), adding machine or difference engine to actually function reliably, or relays and vacuum tubes to last in the face of bugs, involved a lot of security engineering in that sense.

In terms of a specific "security feature" still in use:

  • for data storage and transmission integrity, a "parity track" was present on the first magnetic tape data storage in 1951.

Perhaps some would want to describe the ability for computers to modify their own instructions (Von Neumann architecture) as the oldest "anti-security feature"? It was theorized in 1936 (the Turing machine), and designed in 1944 (EDVAC).

If you buy that, then the use of the Harvard architecture for separation of data and instructions, named after the 1944 Harvard Mark 1, but in use since the beginning of computers, would also be an early "security feature".

Update: taking a cue from @thomas, we can look back before the fabulous ant world to lots of much older mechanisms:

  • 1
    +1 @nealmcb: If you're going the route of ancient astronomical instruments (aka mechanical computational systems, 100-150 BC), the Antikythera mechanism if my favorite example of one. More over, just to give some idea of how complex this system was, here's a time-lapse video showing in Legos how hard such a system would be to replicate functional. Beyond that, I agree computational systems in themselves must provide mechanism that provided integrity and availability. – blunders Jun 6 '11 at 12:45
  • 1
    Getting things to work against the forces of nature and random errors is not security engineering. Security is concerned with intelligent adversaries. – starblue Jun 6 '11 at 14:27
  • @starblue: Is having redundant power sources a security issue? If so, a force of nature is the most common reason that countermeasure would be required. One example of random of errors being security related would be fuzz testing. Moreover, the main issue, integrity of the system - an example, an IDS that logs calibration patterns and issues an alert if modifications are of note might have caught the Stuxnet attack. – blunders Jun 6 '11 at 16:49

I'd agree with @Vineet and say it would be ciphers. Specifically I'd be inclined to point out the Caeser Cipher, so called as it was used by the Roman military to protect messages. May not be the oldest cipher in use but it's the oldest one I'm aware of.

  • Not the oldest, the Bible has several examples of primitive ciphers. – AviD Jun 6 '11 at 9:47
  • @AviD: for that matter, some "secret writings" using non-standard hieroglyphs have been in use in ancient Egypt since the 3rd millenia BC, more than a thousand years before the Bible began to be written at all. – Thomas Pornin Jun 6 '11 at 11:13
  • Aherm, cough cough... @Thomas, I'm really not gonna get into a theological debate with you here ;)... But anyway, the hieroglyphs sounds interesting - do you have more info on that? – AviD Jun 6 '11 at 11:17
  • 1
    @AviD: David Kahn, in "The codebreakers: the story of secret writing", cites an example from 1900 BC ("in the main chamber of the tomb of the nobleman Khnumhotep II"). The "middle of 3rd millenia" date is quoted from the Encyclopaedia Britannica (in the "hieroglyphic writing" article). The cryptographic nature of these non-standard hieroglyphs is somewhat uncertain; for that matter, the whole chronology of ancient Egypt is not completely clear either (depending on authors, dates may vary by up to 300 years...). – Thomas Pornin Jun 6 '11 at 11:51

The "book cipher" ?

I don't know how anyone would be able to authoritatively state which "security feature" was used first in history. But I think the book cipher, or variations on the lookup pad theme such as the one-time pad, is a fair guess.

  • -1: The one-time pad sounds pretty simple to us today, and so it would be a good guess that it's one of the earliest. However, it was first described in 1882; there are many other ciphers that are first described centuries or millenia earlier. – David Cary Aug 9 '12 at 17:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.