In den Flüssen nördlich der Zukunft
werf ich das Netz aus, das du
zögernd beschwerst
mit von Steinem geschriebenen
Schatten

A relatively “simple” Celan poem. Two movements: the speaker casts a net, the addressee loads it with “stone-written shadows.” There is one main clause and one relative clause. Pierre Joris translates it fairly well:

In rivers north of the future
I cast the net, which you
hesitantly weight
with shadows stones
wrote.

He is trying to avoid the passive construction: “shadows written by stones.” In the process he embeds a second relative clause in the first: “shadows [that] stones wrote.” It seems to me that the poem should end with the word “shadows,” and that the introduction of a second relative clause, with a third active verb, breaks up the the poem’s rhythm: “I cast the net, which you weight...” I generally approve of Joris’s translations because they are very literalistic, that is, they attend to the letter of the text. In this case, a departure from literal translation is not necessarily an improvement.

A misplaced comma or two can ruin a translation. Here is Spanish poet José Angel Valente translating the same poem:

En los ríos, al norte del futuro,
tiendo la red que tú
titubeante cargas
de escritura de piedras,
sombras.

Valente intuits that the poem should end with “sombras” (Schatten, shadows), so he places it in apposition to “escritura de piedras,” squelching the idea that the shadows are written by stones. The original had two clauses separated by a comma. Valente introduces three extraneous commas, and omits the one separating the two clauses. The rhythmic effect is destroyed. (In German, as I remember from my four semesters twenty years ago, the comma before the relative clause is obligatory.)

Yet Valente’s translation works quite well if we don’t care about the German original. It has its own, broken, stuttering rhythm. The alliteration of “tú / titubeante” is particularly effective. “Cargar” is a good verb here because it means “loading” or “charging.” A participle (geschriebenen) becomes the noun “escritura,” Spanish for “scripture” as well as for “writing.” I visualize huge stone letters being loaded into a net, letters which the poet associates with shadows, a moment later. “Sombras” could be in apposition to either the entire NP (escritura de piedras) or to the noun (piedras). A stonewriting or a shadowwriting? Here is my version of Valente’s translation:

In rivers, to the north of the future,
I lay down the net that you
clumsily load
with a scripture of stones,
of shadows.

Valente reworks some of these images in another poem that begins “Al norte / de la línea de sombras / donde todo hace agua” (To the north / of the line of shadows / where all becomes water). This poem also contains a concealed reference to Celan’s theory of poetic communication as a “message in a bottle.”