When you like or reblog a post, that counts as user engagement for the person you liked or reblogged from, and shows up in their notifications.  

If the person you liked or reblogged a post from wasn’t the original poster (i.e., you’re liking or reblogging a reblog), it also counts as user engagement for the original poster, and shows up in their notifications as well.  

This means that user engagement from your likes and reblogs can potential accrue to two different people, the original poster and the person you liked or reblogged from.  

Consequently, you cannot “steal” user engagement from someone by reblogging their post.  

This is one of the very few areas where Tumblr is actually functions more reasonably than other social media platforms.  

Note that this is only true if you use Tumblr’s built-in reblogging function. If you save someone else’s content to your local device and append it to a new post, you effectively become the original poster from that point on.  

This means that on Tumblr, “reblogging” and “reposting” are two different things; if you see someone complaining about “reposting”, this is not the same as reblogging.  

Commenting when reblogging does not affect any of this – unlike, say, Twitter, where quote-retweeting causes user engagement to accrue to the quote-retweet and not to the original tweet – and you can and should do so freely.  

However, every Tumblr user can see who exactly you reblogged a post from, which functions as a soft disincentive against making inane comments; if you make a dumb comment on a reblog, people who see your reblog may “back up” one step in the reblog chain to reblog a version of the post without your comment.  

Nobody understands tags, and there’s a fair amount of evidence that how tags work changes periodically and without warning.