While consulting the version of the New Oxford American Dictionary provided with Mac OS, I came across the following “word spectrum”, which its says “show shades of meaning between two polar opposites”. While I generally appreciate its notes such as “The right word”, I was surprised by this spectrum:

Word spectrum

Some of the words in there appear next to one another, which I would truly treat as synonymous, like “honest” and “truthful”. However, I am surprised by the ordering of some other words. For example, I would have classified “upfront” even higher than “frank”. Also, I would classify words like “taciturn” and “of few words” as neutral on this scale, as I don't feel they convey any negative connotation.

So, my question is (and I'm sorry it took so long to get to it!): is this spectrum corroborated by usage in your experience? That may depend on local factors, of course, which I'd be happy to be referred to.

  • For who is looking for the diagram, it is shown from the Dictionary.app thesaurus for frank. A note after the diagram reports that Word Spectrums show shades of meaning between two polar opposites.
    – apaderno
    Feb 20, 2011 at 12:52

3 Answers 3


This is a difficult question to answer, mostly because there are so many words in the spectrum. Trying to draw out some key points:

  • The scale is one-dimensional, and there are multiple dimensions to consider: truthfulness/deception; good/evil intent; pure information content; clarity of information. I believe the scale makes most sense if you look at it as rating the words on "how much information you receive": vague is very close to the bottom, but does not necessarily imply deception or ill intent.

  • That said, of course a lot of the words have connotations besides their position on the scale - simply assigning an "information content percentage" to each one would not tell the full story of their meaning. For instance, sly and slippery have a connotation of wishing to deceive for nefarious purposes: if I avoided answering someone's questions because I didn't want to spoil the surprise of a present they were going to receive, that could certainly be called evasive, but it would be unlikely to be described as sly or slippery; similarly, someone's answer can be "concise" without being "to the point" - if it's brief but doesn't address the issue.

  • With many of these words, there is a difference between applying them to a person about their character in general, and applying them to their response (or otherwise) to a specific question. For instance, saying "John is quiet." is neither positive nor negative - it just means he doesn't tend to make a lot of noise; whereas "John is quiet on this matter." starts to sound as if he may be being deliberately evasive, and hence is slightly negative.

  • The word taciturn falls into this category - applied generally I'd say it's fairly neutral, but applied about a specific topic it implies more evasiveness than quiet. But all of these words appear at the middle of the scale, with mostly neutral words between (except for uncommunicative, which to my mind falls somewhat lower down the scale), so there's not necessarily anything wrong here.

  • I would also place upfront into this category: saying "Bob was upfront about this matter" simply means he made an effort to inform us about it / did not attempt to hide anything; however saying "Bob is an upfront kinda guy" would suggest that he expresses all feelings and opinions forthrightly all the time, even without invitation - which might well qualify for placing above frank.

  • Similarly some of the words apply to the person, and some to their response (the prime example here being short and sweet, which if applied to a person would describe their height and looks rather than how talkative they are!); on the other hand one would rarely say "his answer was silent".

  • On top of all that, I'd say that any attempt like this to rank words along a continuum is always going to be approximate, unless the words have precise, technical definitions that allow it. Therefore the diagram should be taken with a pinch of salt - just because one word is a little higher or lower than another doesn't necessarily imply it is stronger or weaker (or even appropriate) in all contexts.

So, with all of those qualifications (and there are a lot!) - I'd say that the classification is a reasonable guideline, in that each word listed can be used in a way that justifies its position on the scale, though many can also be used in ways that contradict it. But it would be good if it had been published with some of these caveats alongside it :-)


First, let's consider the definitions a priori.

'Frank' is about content and form: if I am frank, I will not hesitate to tell you what I think, and I will say it in a straightforward manner. The majority of the words listed are about form only, content only, or verbosity only: one could be sincere in their content and flowery and verbose in their phrasing while still being of few utterances, which would probably mark them as taciturn.

So, a priori, this spectrum is incorrect, since it goes through at least three axes and is inconsistent about which are required.

Next, let's consider the spectrum as a spectrum reflected by human speech. There are 56 gradations listed. I'm not doubting that humans have a very nuanced view of honesty and forthrightness and verbosity. However, conveying these nuanced views consistently enough for any spectrum of 56 words to be meaningful strains credulity.

Were I forced to put those words into a spectrum, I don't doubt I'd come up with something similar, but those words simply do not fit in the same spectrum.

  • 1
    Yeah, of course, usage of words can never be ranked on scales. However, I am interested in how people think this scale does, in first approximation, to classify the words that are on it.
    – F'x
    Feb 20, 2011 at 13:55

This is a constellation of words that, like star constellations, only looks like something recognizable when seen in two dimensions. In point of fact, terse, concise, innocent and many others in this list (sly? slippery?) are not directly on a path between frank and evasive.

This kind of exercise only shows that human beings see patterns where none exist because our minds are programmed to recognize patterns. I think the continuum would be clearer if it were edited down to fewer than a dozen words. A word cloud that gathers nebulous relationships together and delineates them explicitly does nothing to promote clarity.

  • I think this is an example for trying to impose (not even see) a pattern were none actually exists. As @dhasenan says, it's forcing points from a multi-dimensional (not sure how many, but certainly more than one) space onto a one-dimensional axis. Which doesn't work (just like there's no total ordering of complex numbers, pardon the math) Feb 20, 2011 at 16:48
  • @jae: True, but you can project the complex numbers onto the real/imaginary axis, and order them by that. But then it pays to be very explicit that that's what you're doing, and with which axis - which is what I feel is missing here :)
    – psmears
    Feb 21, 2011 at 22:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.