@Zetherin,
Dictionaries are never a definitive reference of language, to a language, or the like. The best a dictionary can do is provide a relative aggregate of exposition on a given word. See my previous post about paraphrasing.
Likewise, one cannot simply emulate speakers of a language, not successfully at least. In that vein, learning from a dictionary and emulating speakers will always result in contradictory results to the outsider, simply because either individually is inadequate to grasp the essence of the language.
In truth, nothing is sufficient. What is explained as grasping the essence of a language; 'getting it', is as natural as walking or eating, in the sense that once complete it is not a conscious process and furthermore is not confined to any specific manner or form. There are only regions, spheres of influence in which the speaker is well established enough in a linguistic community to provide his own input, while at the same time be perceived as an actual speaker of the language.
Dictionaries make an artificial summaries of all these classes and of all these peculiar usages, meant to serve as a reminder to the actual speaker when placing a given word in a particular context; it should be seen as a crutch, but not an actual pair of legs, and for one who can properly walk, the crutch is awkward, and cumbersome.
But as Gosh said, this is basic linguistics; though it is not the popular discourse among other academic fields, certainly not in the so-called logical fields, it is nevertheless written and explained all over in most introductory texts on linguistics. Simple Aristotelian logic will not work.
Thus a dictionary is equivalent to a literary work; it provides a window as to what people have anticipated words to mean, it provides a perspective for the state of that particular society and culture. Like a monument to a war does not comprise victory itself, and like a glorified statue of a deity does not comprise that deity's actual traits and attributes, so does a dictionary not comprise the actual usage of language. All of these artifacts depict collective anticipations which are to be viewed in a collective sense.
For in essence a dictionary does nothing more than use other words to explain any index. Thus it inherently assumes that the user is already familiar with those words used as explanation, to some extent. Assuming otherwise would be making the entire ideal contradictory, as on the one hand the dictionary would provide the definitive prescriptive reference for language, but yet on the other hand assume that the reader already uses all the descriptive terms according to the dictionary itself; thus making it a horrible instructive tool.
You may see a lot of familiar faces around, Pyrrho, Kennethamy, Fast, Leaf, cberman, and maybe some more I cannot currently recall.
