The Glamour of the Grammar: Dot one, part two

A recent column ended with an admonition about this well-known bit of grammatical advice. Today we'll take a closer look.

Hebrew Hear-Say logo (photo credit:)
Hebrew Hear-Say logo
(photo credit: )
'Beged kefet b'rosh mila," that is "beged kefet [letters get a dot] at the start of a word." A recent column ended with an admonition about this well-known bit of grammatical advice. Today we'll take a closer look. The rule is an attempt to explain when the six letters bet, gimel, daled, kaf, peh and tav get a dagesh (dot in the letter) even when they're not doubled. Though the phrase is etched into the mind of every Israeli high-school student, it's not quite right. In fact, there are two things wrong with it. Before turning to the problems, let's look at the motivation behind this "rule." In the Hebrew word bayit (house), for example, the bet at the beginning of the word has a dagesh and the tav does not. By contrast, in the word teiva (box), the tav at the beginning has a dagesh and the bet does not. According to "beged kefet b'rosh mila," this is because a dot is added to the first letter in each of the words, but not to any other letter. Similarly, in the Hebrew word for "roof," gag, the first gimel but not the second gets a dagesh. So it looks like beged kefet letters at the beginning of a word get a dagesh. But it's not so simple. The first problem with the rule is that it doesn't work. Israel's most popular fast food, falafel, provides a convenient demonstration, because according to the rule, it should be palafel. But of course it's not. And the Hebrew word for "chemistry" is chimya, with no dagesh in the initial kaf. (And when you see something you don't like, you say fehh, not pehh.) So maybe the rule works for some words? For words that come from classical Hebrew? It almost does. But this is the interesting part, because, in fact, the rule is backwards. It's not that the beged kefet letters get a dagesh at the beginning of a word, but rather that they come with a "built-in" dagesh, and lose it after a vowel. So, in our examples from above, the bet in bayit comes with a built-in dagesh. So does the tav, but it loses the dagesh after a vowel. Similarly, the first gimel in gag doesn't change, but the second loses its dagesh. Why does it make a difference? Because there are two stages in the production of a word during which a beged kefet letter might lose its dagesh. When we discussed the kamatz katan, we noted that words have base words (for example katav, "he wrote") and derived forms (for example, katva, "she wrote"). Similarly, lichtov, "to write" (being an infinitive) is a base word, while the seemingly very similar bichtov, "in/by writing," is a derived word, from b- (in/by) and k'tov (writing). After a vowel in a base word, a beged kefet letter loses its dagesh. So, in katav, the tav and bet lose their dagesh, while the kaf does not. Sometimes (for complicated reasons), a base-word vowel drops out in a derived word. Katva is an example. The vowel between the tav and bet drops out. But the vowel de-dagesh-ifies the bet (turning it into a vet) before it drops. More generally, base-word vowels do their dagesh work before they drop out. This is also why, believe it or not, lichtov has a dagesh in the tav (because there's no vowel before the tav in the base word) while bichtov has no dagesh in the tav; the dagesh fell out before the initial bet got added. A vowel that gets introduced in a derived word also forces letters to lose their dagesh. So bayit has a dagesh in the bet, because the bet comes built-in with a dagesh. By contrast, "to a house" is l'vayit, with a vet. (This rule, though, doesn't really apply to Modern Hebrew. "In a cup or in a can" ought to be b'chos o b'fachit, but except for linguistic snobs, it's b'kos o b'pachit.) We're almost done. The final bit of information is that a doubled letter (almost) never loses its dagesh. So, for example, in diber (he spoke), the dagesh in the bet represents a double bet, and therefore the hiriq (/i/ sound) can't get rid of it. So that's the basic story. There are few refinements that will have to wait, but for now, you know enough to understand most dagesh dots. The writer teaches at HUC- JIR in New York City.