I found the reference I mentioned in my previous post. It is Figure 18 in Chapter 12 of Psychology of Intelligence Analysis
The table shows the probability assigned by various reader to statements containing verbal expressions of uncertainty. For some terms, such as 'highly likely' or 'probably' the probability assigned by various users spanned a range as high as 50%. Clearly using verbal descriptions of probability is probably :-) a bad a idea!
Nicolas Bissantz beat me to this topic by at least 1 week in his posting. Speechless not numberless He gives examples from the financial and medical area.