First you need to implement proper pagination canonicals, either as
rel=prev/next or as a "view all" page using a catchall URL. Here is a guide to do that https://support.google.com/webmasters/answer/1663744?hl=en
Keep in mind, if you use a "view all" page, its canonical must be static, meaning you cant just use
/my-category&limit=234 since it would often change. Instead, a static URI such as
/my-category&limit=all works much better.
After you set those up, head on over to Google webmaster tools and visit "Crawl > URL Parameters". This is where you teach Gbot how to use parameters, sorts, limits, filters, or any other querystring that mitigates the page. So you can say to Gbot "Hey, if you see something like
&sort= it means sort the results by X parameter". You can do the same thing for Bingbot although its more of an "ignore this" than a "learn this" situation.
Finally, if you have extra/utility querystrings that arent meant to be used by public users you can disallow them in robots.txt. This is normally not recommended unless there is something like API endpoints or special strings for the URL to turn it into a JSON response. But be careful what you disallow, it could very well bite you in the butt later.