In OP's examples, then means next, [immediately] after [whatever time/event was previously mentioned].
The concept of transitional tags seems useful here. Depending on your point of view, you could say then conveys either addition or time sequence - but I'll go for the latter, since that's obviously the category for related alternatives such as first, next, finally.
My initial inclination was to say then can happily go before or after the subject (I in OP's example), and leave it at that. But it's worth pointing out that although this is true, and applies equally to first and next, it doesn't apply to finally in quite the same way...
1a: First I did this. 1b: Then/next I did that. 1c: Finally I did something else.
2a: I first did this. 2b: I then/next did that. 2c: ?I finally did something else.
There are some contexts where (2c) above is acceptable, but it doesn't really work in contexts where the only sense of finally is "next and last". If finally comes after the subject, there's invariably a strong implication of "next, and at long last".
I think what this suggests is that all these "transitional tags" should normally come first in the statement, but by custom and practice native speakers don't mind moving small short words to between the subject and verb. But because that's slightly "unusual" positioning, we tend to look for a possible semantic difference.
In practice there doesn't seem to be any credible semantic difference depending on whether we place then before or between subject/verb, so it's just a stylistic choice. But there certainly can be a subtle difference according to the position of then...
3: "You lost £1000 at the roulette table? I hope you then stopped gambling!"
4: "You lost £1000 at the roulette table? I hope you stopped gambling then!"
In (3), then has the normal sense (chronologically after), but in (4) it has the more "extended" sense of as a result [of what came before], in that case, therefore.
To summarise: there's no grammatic or semantic difference in OP's specific examples, but if you're looking for general principles, you'll be safer if you put any transitional tag first in a statement.
Finally, here's an NGram showing how usage has shifted over the past couple of centuries.
Compare the following sentences:
Present time
- It's very dangerous in Metropolis right now. You can be beaten up for just looking at someone the wrong way.
- It's very dangerous in Metropolis right now. You could be beaten up for just looking at someone the wrong way.
Sentence (1) presents being beaten up as a live possibility, as something that does actually happen to people. Notice that we can analyse for just looking at someone the wrong way, as some kind of condition:
1'. You can be beaten up if you just look at someone the wrong way.
Sentence (2) presents being beaten up as a hypothetical outcome of the wrong look. It is not being presented in the same way as in (1). It doesn't necessarily say that people are being beaten up for looking at people the wrong way - although we might assume that they are.
Could in sentence (2), as already mentioned, represents this outcome as a hypothetical possibility, as opposed to a live one. We could rephrase it as the following hypothetical conditional:
2'. You could be beaten up if you just look(ed) at someone the wrong way.
Remember both of these sentences refer to the present or future time. The way that we interpret you here is quite likely to affect our interpretation of the sentence. It's also quite likely to affect our choice of can or could. If you means a person in general, we are more likely to use can. If you is being used to make the listener imagine themselves in that situation, then we are more likely to use the hypothetical could.
Past time
Let's move forward twenty years. Now if we wish to make the same kind of statement but about Metropolis twenty years ago instead of now, we need to shift the tenses back to indicate past time. Example (1) would now be as in (3):
- It was very dangerous in Metropolis in those days. You could be beaten up
for just looking at someone the wrong way.
Here we see could appearing as the past form of can. This past form of can still implies that people actually were being beaten up for looking at people the wrong way. It is being presented as a live possibility for people at the time. It has exactly the same meaning as can, but refers to a past time. It does not represent hypotheticality.
However, in sentence (2) we already have past tense could, where the past tense indicates hypotheticality. If we wish to keep this hypothetical flavour, but also indicate past time, then we need to shift the tense back further. We need to use a past perfect form as in (4):
- It was very dangerous in Metropolis in those days. You could have been beaten up
for just looking at someone the wrong way.
Sentences (3) and (4) could be construed as the conditionals:
3'. You could be beaten up if you just looked at someone the wrong way.
4'. You could have been beaten up if you ('d) just looked at someone the wrong way.
So if we see a past perfect form could have this most likely represents past time reference plus hypotheticality.
Hope this helps!
Best Answer
Your first clause is using inversion without the word if. You can only do that to form conditionals if you use subjunctive inversion, which you are not doing here.
Only the last version there permits you to forgo the if by virtue of using subjunctive inversion.
There are many other ways to form conditionals, but what you have written is not one of them.