Our Magnificent Bastard Tongue (20 page)

Today, Semitic’s most prominent representatives are Arabic and Hebrew. In the last centuries B.C., however, these were both obscure languages of small groups, and the shop-window Semitic representatives, used as lingua francas in the Middle East and/or beyond, were other ones. Akkadian is often mentioned via the names of its dialects Assyrian and Babylonian. Aramaic was once so entrenched as the language of note in the Middle East and beyond that it was the language of administration under the Persian Empire, run in Persia vastly eastward of where Aramaic had arisen, despite the native language of Persia’s rulers being, well, Persian, completely unrelated to Aramaic. It lives on today among small groups, termed, for one, Syriac. Akkadian had
z
,
s, sh, ts,
and an
h
sound that you made with your uvula. Aramaic at the time had
sh
,
dz, ts,
and
h
. Snaky sounds.

So, just hypothetically, if speakers of languages like these wrapped their tongues around Proto-Germanic, we might expect that their rendition would have more hissy sounds than Proto-Indo-European passed down to it. But this alone can be so compelling only as a speculation. For one thing, one other Indo-European branch went hissy, too, apparently all by itself: Armenian, which occupies its branch all alone.
Pater
(
father
) in Latin,
hayr
in Armenian.
Cor
(
heart
) in Latin,
sirt
in Armenian.

But there was something else about Proto-Germanic.

To an English speaker it feels pretty normal that as often as not, we put a verb into the past by changing the vowel in it instead of adding -
ed
:
see, saw; drink, drank; come, came;
etc. And in Germanic in general, it is indeed normal: in German, those verbs are
sehe
,
sah; trinke, trank; komme, kam.
But in Indo-European, beyond Germanic, this is not normal at all.

You may know this from taking French or Spanish: there are certainly irregular verbs, but the irregularity is only rarely just a matter of switching a vowel. In Spanish, you start with an innocent infinitive form like
tener
(“to have”), and then cut your teeth on mastering that
he has
is
él tiene
but
he had
is
él tuvo
. It’s not just the
u
vowel—there is also that random
v
that comes out of nowhere. Typical—and not just a matter of vowel switches alone, like
come
,
came; drink, drank.
These, where it’s all about the one vowel, are Germanic’s kink. In all Germanic languages, there is a long list of verbs whose pasts are formed like this, traditionally termed “strong verbs.”

The reason this is not the case in other Indo-European subfamilies is because Proto-Indo-European was not like this. Its grammar did involve switching vowels—but to do an array of things such as helping to indicate case: if you asked a Proto-Indo-European speaker what a dog was called, they would have said it was a
kwōn
, with a long
o
. But in the genitive it was
k
u
n-és
with a
u
, and in the accusative,
kwón-ṃ
with a short
o
instead of a long one. Indicating past tense was only one thing vowel switching was used for (
know
,
knew
was
wid
,
woid
)—and only so much. In other branches of Proto-Indo-European, this vowel-switching machinery was passed down in assorted renditions reflecting that array of functions it had in Proto-Indo-European. Only in Proto-Germanic did the all-over-the-place vowel-switching of Proto-Indo-European morph into something as distinct and particular as a long list of past tense verb forms indicated with a vowel change and just that.

Once again, Proto-Germanic is odd. That’s in two ways now. Might there be a reason? Well, what about those Semitic languages again? Interesting—their kink is that they form the past tense by changing the vowels inside the word. In Hebrew today,
he writes: hu k
o
t
e
v.
But
he wrote
:
hu k
a
t
a
v.
The consonants stay the same: the vowels change:
write, wrote; kotev, katav.
All Semitic languages have had this feature, ancient and modern, including good old Akkadian and Aramaic. Hmm.

Even these two things are not quite a smoking gun, but there’s something else.

Amid early offshoots of Proto-Indo-European, Proto-Germanic was like English amid Germanic languages in how much frippery it had lost. Proto-Indo-European marked its nouns with eight cases. Latin, the early Indo-European language most learned in modern times, held on to six: nominative, genitive (“of the table”), dative (“to the table”), accusative (table as object), ablative (“by the table”), and vocative (if you were moved to say “Oh, table!!” but more usually, of course, with names), and then some words had a locative (
Romae
, “in Rome”). The ancestor of today’s Slavic languages, Old Church Slavonic, had seven cases, as Lithuanian still does. Old Irish, an early Celtic language, had five, like Ancient Greek then and Albanian now.

But Proto-Germanic had just four. Those four cases in German wear out Anglophone learners today, but in the grand scheme of things, they are a broken-down half of what Proto-Indo-European had.

In the same way, as Indo-European languages go, it’s weird that in English the only verb endings are ones for present and past tense.
I wait, I waited,
and that’s about it. German, as busier as it seems to us Anglophones, is pretty much the same:
ich warte, ich wartete.
There are no endings that mark the future, for example: English does future with a word,
will;
German uses its
werden
in the same way. That’s how it is in all Germanic languages; that’s how it was in Proto-Germanic.

Note, however, that in Spanish, you use endings to mark not only the present (
yo hablo
) and past (
yo hablé
), but imperfect (
hablaba
), future (
hablaré
), conditional (
hablaría
), subjunctive (
hable
), and imperfect subjunctive (
hablase
). Spanish is not unique here, but normal: it has stayed like Proto-Indo-European was, in which there were separate tables of endings to place things in time according to very specific gradations almost imposingly baroque. In fact, Spanish has taken this even further than Proto-Indo-European did in some ways, sprouting its own new endings. Proto-Indo-European, for example, did not have conditional endings.

This anality about assembling sentences very precisely regarding time and hypotheticality with endings was already de rigueur way back when Spanish was Latin. While Latin was spoken, when Proto-Germanic endings were down to just marking whether something was happening now or already had, Latin endings were painting a much more particular picture of how one experienced actions in time:
portō
(“I carry”),
portābam
(“I was carrying”),
portāvi
(“I carried”),
portāveram
(“I had carried”),
portābo
(“I will carry”),
portāverō
(“I will have carried”).

Where did all of this go in Proto-Germanic? Some descendants of Proto-Indo-European have held on to more of this stuff than others, but in Proto-Germanic it fell away to a peculiar extent, such that we Germanic speakers have dragooned little words like
will
and
would
to pick up the slack.

After what we have seen in this book, the reader will intuit that this suggests that Proto-Germanic was not just bastardized by some other language, but beaten up by it. The streamlining of Proto-Germanic, with its four little cases, and suffixes marking just two little tenses, is the sign of busy adults making their way in the language as best they could but never quite mastering the subtleties. Proto-Germanic seems to have been a kind of schoolboy Proto-Indo-European.
¹⁰

At this point, many will see it as at least worth asking: just who might these people stirring up Proto-Germanic have been?

A final and conclusive piece of evidence that there were, at the very least, some people of some kind stirring things up is that no less than a third of the Proto-Germanic vocabulary does not trace back to Proto-Indo-European.

With the other two-thirds, we can first figure out what the Proto-Germanic word was, like
daukhtrô
for
daughter
, and then we can compare that word to
daughter
words in the other Indo-European subfamilies, and work out that the Proto-Indo-European source root was
dhugəter
.

But with a mysterious many of the Proto-Germanic words, we just hit a wall. There are no cognates of these words in other Indo-European languages, and thus no ancestral Proto-Indo-European word can be reconstructed. Earlier than Proto-Germanic the trail runs cold. The words quite often refer to seafaring (
sea
,
ship, strand, sail
), war-making (
sword
), fish (
carp
,
eel
), and formal social institutions (
knight
). Note, for example, that there is no word akin to
sea
in any other European language you might be familiar with. In Romance, it’s words like French’s
mer
and Italian’s
mare
. In the Slavic languages, Russian has
more
, Polish
morze
. Over in Celtic, in Welsh the word is
môr
. From the shape of all those words, it is no surprise that it is thought that in Proto-Indo-European there was a word
mere
that referred, at least, to something like a lake. But in English we instead have this
sea
thing, with cognates like German’s
See
and Dutch’s
zee
. Why don’t we English speakers refer to eating something like “Mar-food” instead of
seafood
? Sure, Germans, for example, also have
Meer
as an alternate. But where did their
See
come from?

Now, one way of approaching this is to just treat it as an accident. Who knows, after all, whether in other Indo-European languages, words related to the seemingly orphan ones in Proto-Germanic were once alive, but happened to drop out of use by chance? Maybe it happened that a Proto-Indo-European root now survives only in one of the many branches of the family. Although we would have to wonder why it would have blown away in so many branches so uniformly—but still.

Or, the meanings of words can change so much over time that a Proto-Germanic word could possibly trace to a Proto-Indo-European root with a completely different meaning, such that no one would suspect the connection. The word
punch
, when referring to a drink, comes not from a Proto-Indo-European word for some kind of liquid, but from its word for
five
—through Hindi’s
five
word, used because the original recipe was developed in India and used five ingredients. Shitte happens.

And besides, when scholars put their heads to it they can often figure out a Proto-Indo-European root that the Proto-Germanic words could have come from, via likenesses that no one happened to notice before.

So one leading scholar of how languages change has only this to say about the issue in a recent work:

 

Well, yes. But what about when the mysterious words look—mysteriously—like ones in other language families? Like, say, Semitic?

For example, one of those words that does not trace before Proto-Germanic is
fright
. Its spelling reflects that there was once an extra consonant sound before the final
t
, and its rendition in other Germanic languages often gives us a better sense of the original, such as German’s
Furcht
, pronounced “foorkht.” The extra consonant was the sound of
ch
in
Bach
: the Proto-Germanic form was
furkhtaz
.

But check this out. The Proto-Semitic verb for “to fear,” as it happens, had the consonants
p
-
r
-
kh
. I give no vowels because Semitic verbs
are
their trios of consonants; the vowels change to mark tense and other distinctions, as in
kotev
/
katav
(“write/wrote”) above, and there is no “default” vowel pattern that means nothing and signifies a “basic” form. Thus, the closest we can come to what “the word” for
fear
in Proto-Semitic was is
p
-
r
-
kh
.

Now, we can compare
p
-
r
-
kh
with the consonants in
furkhtaz
:

 

The
f
and the
p
don’t look related at first, but then remember that in Proto-Germanic,
p
got turned into
f
! The
p
-
r
-
kh
root for “to fear” just might have also ended up as the word
fright
in England, and hence, on this page.

Or,
folk
started in Germanic as a word referring to a division of an army, and only later morphed into meaning a tribe or a nation. The Proto-Germanic word was
fulka
; the early Semitic root for
divide
—i.e., as in making a
division
—was
p
-
l
-
kh
: