←back to thread

Gettiers in software engineering (2019)

(jsomers.net)
303 points FigurativeVoid | 4 comments | 14 Oct 24 18:28 UTC | HN request time: 0.694s | source
Show context
rhelz ◴[14 Oct 24 21:48 UTC] No.41842371[source]▶
The impossibility of solving the Gettier problem meshes nicely with the recent trend to Baysianism and Pragmatism. Instead of holding out for justified true belief and "Bang-Bang" either labeling them True or False, give them degrees of belief which are most useful for prediction and control.
replies(1): >>41844870 #
1. ars ◴[15 Oct 24 04:10 UTC] No.41844870[source]▶
I don't understand the Gettier problem. The example of the cow for example: You do not have a justified belief there is a cow there, all you can justify is that there is the likeness of a cow there.

To be able to claim there is a cow there requires additional evidence.

replies(2): >>41847683 #>>41852829 #
2. mistermann ◴[15 Oct 24 12:08 UTC] No.41847683[source]▶
> You do not have a justified belief there is a cow there, all you can justify is that there is the likeness of a cow there.

Is this assertion not self-refuting though?

3. rhelz ◴[15 Oct 24 20:35 UTC] No.41852829[source]▶
The cow example is a confusing example; I like the clock example much better. You are in a school building, with hundreds of classrooms, and there is a clock on each wall. All of the clocks are working perfectly, except for one classroom where the clock is stuck at 2:02.

Every other time you've been in that school building, the clocks have shown you the right time, so you feel very confident that the clocks on the wall are accurate.

But this time, you happen to be in the room with the non-functioning clock. It says "2:02" but by great good fortune, it actually happens to be 2:02.

So your belief is:

1. True. It actually is 2:02.

2. Justified. The vast majority of the time, if you see a clock on a wall in that building, it is working fine.

But should we say that you know the time is 2:02? Can you get knowledge of the time from a broken clock? Of course not. You just got lucky.

In order to count as knowledge, it has to be justified in the right way, which, alas, nobody has been able to specify exactly what that way should be. So far, nobody has come up with criteria which we can't find break in a similar way.

// all you can justify is that there is the likeness of a cow there //

If you see something which looks real, you are justified in believing it is real. If you see your friend walking into the room, sure, you've seen your friend's likeness in the room. But you are justified in believing your friend is in the room.

So if you see something that looks like a cow in a field, you are justified in believing there is a cow in a field, even though looks may be deceiving.

replies(1): >>41855814 #
4. ars ◴[16 Oct 24 05:10 UTC] No.41855814[source]▶
> In order to count as knowledge, it has to be justified in the right way, which, alas, nobody has been able to specify exactly what that way should be.

First of all you have to be able to test your knowledge, you would test that the clock is correct for every minute of the day. If you missed any minutes then your knowledge is incomplete, you instead have probable knowledge, (using the same methods that physics uses to decide if an experimental result is real, you can assign a probability that the clock is correct).

Also, since when is knowledge absolute? You can never be completely certain about any knowledge, you can only assign (or try to assign) a probability that you know something, and testing your belief greatly increases the probability.

(PS. Thank you for the reply.)