"Behold! A Man!"

Depending on what one means by a specific arrangement of words, the accuracy of a statement is there or not--it is not about how it is received by an external audience, but about whether the concept intended to be articulated is actually true.  One can only define words with other words, as anything else would be the things themselves which words are supposed to convey.  One must create or learn words since they are mere mental constructs.  Definitions are, as constructs that are not the things they describe, of course limited.  Again, it is impossible to provide definitions without using words to define words, and then those additional words would have to be defined, and so on.  Linguistic communication is inherently imperfect because it does not actually connect minds.

It is also still possible for imperfect (on one level or another) or partially ambiguous (and all language involves some ambiguity except for how one's own words and their meaning can be known to oneself) definitions to be valid in a sense.  Such examples are unnecessary, but a story of the rivalry between Plato and Diogenes pertains to how definitions can be incomplete but still ultimately accurate.  Plato reportedly proposed that the linguistic definition of a man is a featherless biped.  After hearing Plato's definition of a man, the cynic Diogenes plucks a chicken, presents it before Plato's Academy, and says some variation of "Behold!  A man!"  This is often recounted as a humorous event, but it is nonetheless not true that Plato was wrong on this matter.

Humans are featherless bipeds, having no feathers and walking on two legs under normal biological circumstances--which does not mean people cannot crawl or that all humans even have legs to begin with--so the definition is correct, though it would also apply to other creatures as well.  Yes, a plucked chicken has no feathers and walks on two legs.  So would a plucked emu or penguin.  This does not make any of them human, and it does not mean that there cannot be other featherless bipeds that are logically possible or that walk the planet.  Plato's definition of a man is not as precise as it could be.  It is not that the statement itself is false.  Yes, many people are poor communicators or are far less precise than Plato in this story, but this does not necessarily mean the idea in question is false.

Two of the errors a person could fall into regarding definitions are the following: there is the error of being intentionally or carelessly vague/inaccurate without correcting oneself and that of thinking that any incomplete definition is automatically invalid.  If someone defined humans as animals, they would not be wrong, since animals are biological creatures and humans are biological creatures, but this is broader than Plato's summary of a person as a featherless biped.  One could more specifically define a human beyond the generalities of a "biological creature" and the more narrow but still somewhat ambiguous summary of a "featherless biped."  All words are arbitrary and vague, though.  This is not solely a characteristic of any single definition.