Most and some are not the same. Most is more than half, whereas some is just any amount.

This may be true, but I'm having trouble finding any LSAT questions that rely on the distinction.

You could conceivably create one. It's unlikely that one will appear, however.

Here are two recent questions that require the understanding that two different majorities of the same set must overlap:

Dec. 2002, Preptest 39, § 2, # 21

Oct. 2003, Preptest 41, § 3, # 25

Great, that's exactly what I needed. Thanks.

So...

**Some** = at least one, many, all.

**Most** = at least half, all.

Take the example:

_____ flowers are red and _____ flowers are tall.

**Some-Some:** No concrete conclusions. However, at least one flower MIGHT be both red and tall.

Most-Some: Same

**Some-Most:** Same.

**Most-Most:** At least one flower must be both red and tall.

I think your last one (**Most-Most**) isn't correct. Example:

Most roses are red: Let's say there are 100 roses. At least 51 are red.

Most red flowers are beautiful: Let's say there are 1000 red flowers. At least 501 of them are beautiful.

Can we conclude that at least one red rose is beautiful? Not really.

At least, that's what came to mind. If this is wrong please let me know.

Yeah most-most relationships don't always yield an inference. The example you just stated is one such case.

Most A's are B's and most B's are C's do not yield any inferences.

But there is a case where most-most does yield an inference. When you have something like Most A's are B's and most A's are also C's. Since at least 51% of A's are B's and at least 51% of A's are also C's, there has to be some kind of overlap between B's and C's, yielding the inference some B's are C's. The key here is a given term (A), most of whose members have both qualities B and C.

graphically: B<--most--A--most-->C

yields: B<--some-->C

you want the term in the middle to have 'most' arrows leading away from it

So back to the original example, something like most flowers are red and most flowers are tall gives you the inference that at least one thing that is red is also tall (the flower that happens to overlap). The key here is that there is a given term (flower), most of whose members have both qualities of being red and tall.

graphically: red <--most-- flowers --most--> tall

yields: red <--some--> tall

If you had something like most flowers are red and most things that are red are also tall, you get nothing because there is no single term in the middle that has two different qualities.

graphically: flowers --most--> red --most--> tall

yields nothing, since the term in the middle doesn't have two 'most' arrows leading away from it.