For practical application see:
The Indeterminacy of Consensus:
masking ambiguity and vagueness in decision
Consider the following:
S1: Flying planes can be dangerous.
Is S1 an ambiguous string of ASCII code? No. Every character is well-defined.
Is S1 ambiguous as a response to the question posed in English,"What kind of planes can be dangerous?"? No.
Is S1 ambiguous as a response to the question, "What activity involving planes can be dangerous?"?
What the notion of ambiguity feeds on seems to be the presentation of an apparently specified characterization, e.g. string of ASCII code, response to a question, in a partially specified context.
S1, characterized as an English sentence -- this characterization is believed by some to be well defined -- , is ambiguous. Perhaps it would be better to say, "multipurpose" and then to specify the context of activity in which the relevant purposes are usually encountered. The string of characters following "S2:"
S2: John is the man to watch
if taken as an English sentence is similarly multipurpose and can serve as an answer to either of the questions
For other purposes, other items, characterized as English words, are multipurpose, e.g. "I" for speaker self-reference, "that" for "distant" reference, "stop" for calling a halt or specifying a traffic sign, etc.
a. Which man should watch? or,
b. Which man should be watched?
Characterizing ambiguous words as multipurpose highlights the important fact that it is human purposes that play the central role in what is sometimes abstractly characterized as "meaning" and then reified as some mysterious object of investigation. But it does not follow that what I mean by a word exhausts what a word means, i.e. what the traditions of practice or the habitual contexts of purpose are in which the word is useful. Nor do I ward off challenges to my use of a word by invoking "what I mean by it". A challenge -- if understood as a query as to the usefulness of my linguistic practice in an assumed context of common endeavor, rather than the invoking of some supposed authority --invites clarification of the relationship of the user's purposes to that of the challenger.
"Meanings" are communal, not idiosyncratic. To think the contrary is to risk an operationally stultifying solipcism -- for it is in contrast to "received" or "traditional" usage that individual variance is made out. (Wittgenstein's parallel point is that in order to doubt, one must know something.)
If B.F. Skinner, say, insists that learning only occurs when some kind of reward is bestowed, our inquiry into the practices that link the terms "learning" and "reward" cannot be dismissed as perverse rejection of his operationally defined terms. What we want to know is whether his usage enlightens the practices in which the terms, "learning" and "reward", have been traditionally useful.
Where the activities undertaken mostly involved talking and commenting upon discourse, e.g. theorizing, philosophizing, many items, presumed well-characterized although wrenched from ordinary contexts of usage, say, as words, sentences, propositions, utterances and the like, lose the univocality a specific context might give them. They become ambiguous or vague. 
Words or sentences appear ambiguous when their context of usage --"communal expectations of purpose" -- is unclear or insufficiently specified to the person to whom they appear ambiguous. Terms appear vague when speaker and listener do not share the criteria by which judgments of contrast and comparison might be made with respect to the "vague" terms. (I do not mean to insinuate by using the word "appear" that terms may "be" vague as opposed to their "appearing" vague. This distinction between being and appearance which is useful in other contexts, say, being ill as opposed to appearing ill, does not seem to offer much in the way of a useful distinction in our present discussion.)
The Reverend Jerry Falwell and many of his followers do not, I would imagine, find the word "perversion" to be vague. And within the framework of the activities they use it to coordinate, it is not, e.g. such activities as proclaiming homosexuals undesirable, movies as obscene, sexual practices as worthy of condemnation.
To say that a term is not vague is not to say that any two persons believing it so, i.e. not vague, would use it in all imaginable circumstances in the same manner to aid their judgment in action. Nor need we presume some kind of isomorphism in their brain states. All that is needed is compatibility of action in those undertakings in which the term is employed.
One might argue that the term "concept" is vague. To people unpracticed in conceptual analysis, the word seems to be a pompous substitute for "idea" and the possibility of comparing ideas held by different persons must seem to necessitate telepathic powers. If one accepts as a starting point a recasting of the term "concept" as "criteria of judgment" and then learns some activities by which persons can be shown to at first not, and then later come to share criteria of judgement, the term "concept" begins to be less vague in such expressions as "a psychological concept of behavior" as contrasted with, say, "a moral concept of conduct".
As with ambiguity, individual purposes, especially when secret, may introduce an exploitable vagueness of concept into a situation. This need not force us to forego talk of concepts, or to conflates them with terms, or words or utterances or the like. It is sometimes highly useful to be able to say, "They are using the same term for very different concepts". Or, "They have two different terms for the same concept." The possibility of making such distinctions rests on our ability to characterize certain situations or activities as expectable, normal or standard. Such a possibility eludes us if we relativize the notion of vagueness too tightly to considerations of individual, idiosyncratic purpose.
 Increased ambiguity tends to increase false consensus in a group. See T. Gilovich "Differential Construal and the False Consensus Effect" Journal of Personality and Social Psychology. 1990, Vol. 59, No. 4, 623-634.
 See, for example, "Pluralism and Criteria: Minimizing Politicization in Public Service Decision-Making" at http://www.newfoundations.com/EGR/PluralCrit.html Also see, "Teaching Case Analysis to Achieve Philosophical Consensus" at http://www.newfoundations.com/EGR/CaseAnalysis.html.