How does a k turn into an s?

I once took a linguistics course and I was amazed at how hard I found it to follow the changes in sounds even though I was really interested in the subject. Usually when I am interested in something, it isn't hard for me to understand it.

Now I am reading The Horse, the Wheel and Language, which I am enjoying even though the linguistics part of things isn't easy for me to get through.

The author provides an explanation of the changes for the word "hundred" and I can follow most of it, but I cannot understand how the k sound changed to the s , ie. *k'mtom- to simtas.

I have been k-ing and s-ing for a few days now but I can't follow that change! Can anybody help?

Simply put, the change from [k] > [s] goes via an intermediate stage of [sh] or even [tsh]; there is no direct pathway from [k] to [s] -- the sounds are just too different. If you pronounce a [k] sloppily, at one point, it turns into a palatal fricative ([sh] or [tsh]), and it ceases to be a plosive. From [sh] to [s] is only a (relatively) small step. So the shift from [k] to [s] involves, broadly speaking, two changes (from [k] to [tsh], and from [tsh] to [s]).

Of course, people don't just switch from [k] to [tsh] or from [tsh] to [s] overnight: these changes take centuries to complete. And these changes don't just happen to any [k], either. Usually, the [k]s that change are found in palatal or frontal environments, i.e. next to sounds that are pronounced medially or frontally, such as [i]. When a [k], which is pronounced at the back of the mouth, occurs around these more frontal vowels, its pronunciation will be affected by them.

Would you like some examples? In Old English, some [k]s were palatalised to [tsh]. This sound change underlies the difference between English church on the one hand, and Scottish kirk, German Kirche and Dutch kerk on the other. All of these go back to a Greek word kyriake. The English word birch has cognates in Dutch berk and Swedish björk. Or take the English word I, for instance. The proto-germanic word *ekan "I" developed into Anglo-Saxon ic, pronounced "eetsh". At a later stage, the [tsh] was dropped altogether, eventually leading to today's "I". In the case of German ich, the original [k] developed into [X], a sound similar to Scottish loch. Dutch retains the [k]: the Dutch word for "I" is ik.

People don't pronounce each sound separately: while pronouncing the initial [k] in, say, West-Germanic kirika "church", people are already preparing for the upcoming [i] ; and when they are moving on to the medial [k], they're still halfway through the [i]. This has the effect of pulling the [k] forward, as it were: the pronunciation of these [k]s will be slightly different from those occurring around back vowels, like [u] or [o], leading to a "weaker" plosive, eventually palatalising the [k] to [tsh].

(edited to adjust some characters)

Note on Germanic "k" in words like "ich": The pronunciation of "ich" varies with dialect within Germany. In places it is almost "sh", but in the northern (Platt) dialects, it's still "k". The change is not purely north/south, but pretty much that.

I have a slight addition to Petroglyph's excellent explanation: s corresponding to k in two related current languages might both derive from some other sound (such as [sh] or [tsh] or [X]) in an ancestor language.

My thanks to all of you. That was indeed an excellent explanation, Petroglyph, and I am going to print it and keep it with my notes. Now my husband and I are ichshing and almost-k-ing and using the church and kirk example makes it easier to see (or hear or feel). I can see that these things would make nice puzzles to work on.

When I read books such as The Horse, the Wheel and Language or other sorts of history books, I feel that I am putting pieces into a planet-sized puzzle.

Thanks again!