Sometimes I get the impression from students, and also sometimes from certain scholars and publishers, that reference works are treated as especially trustworthy because they are “objective.” For example, students (and even scholars) will appeal to a Greek lexicon as if it is based purely on facts and figures and no human interpretive element is involved. The same could be said with dictionaries and grammars.
Put another way, we might feel the need to defend why we are working from a monograph, but not why we are taking a statement in a grammar or lexicon at face value. I like to remind my students – who writes these entries? Robots? Computers? No, they are written by people, people with opinions. That doesn’t mean those people are plotting to take over the world, but it does mean there is the possibility of bias and the possibility of oversights and mistakes. (I won’t get into anti-Semitism in the famous Theological Dictionary of the New Testament, but see here).
Advice: read all your sources critically, don’t assume the source gets the information “right” just because it comes from a grammar, dictionary, or lexicon. When you are citing a reference work, don’t just say – TDNT says “such and such.” You are best off actually engaging with the evidence put forth in these articles/entries, not just the conclusions or opinions.
I pretty regularly disagree with glosses and assumptions in BDAG, and I think Louw-Nida is stronger overall in giving accurate meanings and recognizing the polysemous nature of certain words. But this problem is bigger than lexicons. Reference works are tremendously helpful and I use them all the time in research for lectures and scholarship. But the bottom line is – don’t take off your critical lenses just because the cover says “dictionary”!