It's not setting a bound at all. It -is- saying that the amount of time is "surprisingly little". But to the topic at hand, the logical implication is that "to

It's not setting a bound at all. It -is- saying that the amount of time is "surprisingly little". But to the topic at hand, the logical implication is that "to do it" => "takes surprisingly little time" (that is, doing it implies you took surprisingly little time) Not "taking surprisingly little time" => "to do it" (that is, taking surprisingly little time implies you do it; after all, you might have taken surprisingly little time to do -something else-). Same as the above; the implication is that "success" implies "(at least) a surprisingly small number of adults were involved", not that "(at least) a surprisingly small number of adults were involved" implies "success". I find it helpful to use formal logic to explain stuff like this. If I make a statement that p => q (that is, p implies q), and that statement is true, it does NOT necessarily follow that q => p (that is, q implies p) is true (it may or may not be; it is independent from a formal logic perspective). It is necessary to evaluate q => p separately, in which case we may find it to be true, OR we may find q =/> p to be true (that is, q does not imply p; or put another way, the fact q is true tells us nothing about whether p is). As an example, "My shoes are wet because it is raining", p = it is raining, and q = my shoes are wet. p => q means if it is raining (p), my shoes are wet (q). But it does not also follow from that that q => p; that is, just because my shoes are wet (q) it does not mean it must be raining (p; I might have, after all, been out watering the lawn). Thus, q =/> p. For the tango example, p = a tango is possible, q = there are two people. p => q, but q =/> q (that is, two people does not imply a tango is possible). In this case, p = foster child is successful, and q = a number of adults (sometimes a surprisingly small number) were involved with them. p => q, but q =/> p.