Brachos 12a - Mitzvos Require Intent

The gemara discusses a situation where one begins a bracha with the wrong intent, but then finishes off the bracha properly, whether it is a valid bracha. Talmidei Rabbeinu Yona assume that this is dependent on the issue of whether mitzvos need intent. Since the person said all the words correctly, just that he had the wrong intent when he started the bracha, if we assume mitzvos don't require kavan he would still be yotzei. Based on this assumption, Talmidei Rabbeinu Yona asks why the gemara doesn't explicitly answer the question based on other sources that we have regarding mitzvos needing intent, such as the gemara in rosh hashana 28a. TRY (talmidei rabbeinu yona) write that according to the Ri"f who holds that mitzvos require intent, the gemara's question is legitimate. It seems that they understand that even if mitzvos would normally require intent, a bracha may not require intent at the beginning of the bracha since it was finished properly. However, according to those who pasken that mitzvos don't need intent, surely here the bracha would be valid even with the wrong intent at the beginning. What then is the gemara's quetion? TRY answers that when one does an act of a mitzvah such as blowing shofar or shaking lulav, we can pasken that mitzvos don't need intent because the act is significant enough to make up for the lack of intent. However, mitzvos that are dependent on speech, such as a bracha, would surely require intent. The rationale is:

שהאמירא היא בלב, וכשאינו מכוין באמירה ואינו עושה מעשה נמצא כמי שלא עשה שום דבר מהמצות

Since speech is ultimately dependent on one's heart, speech without intent is worthless. That is why the gemara suggests that one could not be yotzei their bracha when they had invalid intent at the beginning.

The Bach (hagahos on TRY) points out that the gemara at the very beginning of the next perek discusses mitzvos tzrichos kavana in the context of saying Shema. That is clearly a mitzvah done with speech, without action, yet the gemara seems to raise the standard issue of whether מצות צריכות כוונה? The answer of the Bach that TRY just mean that the gemara doesn't answer the question here, but does answer it at the beginning of the next perek by showing that this distinction isn't true, is a difficult answer.

To answer this question we need to understand what TRY says. What do they mean that speech is dependent on the heart and therefore intent is integral? There are two types of mitzvos that involve speech. One type of mitzvah is a mitzvah that is really dependent on the saying of something, it is a mitzvah that is fulfilled through אמירה such as shema and kiddush. Another type of mitzvah is where the focus is really on what is in one's heart but needs to be formally articulated to substantiate what is in one's heart. Brachos on food would seem to be the latter category. The purpose of the bracha is to feel appreciation for what Hashem has given, just that the feeling of appreciation needs to be articulated. When it comes to mitzvos that are truly dependent on speech, TRY would categorize them as all other mitzvos that are dependent on action. The speech in shema is tantamount to the action of shaking lulav, where even without intent one can be yotzei. But when it comes to brachos where the focus is really on what is in one's heart, they need to feel the appreciation prior to articulating it. That is what TRY means that it is dependent on one's heart, and that is why intent is more essential for brachos than it is for krias shema.