i wonder if ANY text analysis can be completely free from some external
pressupositions. for example in gen 13:1
וַיַּעַל אַבְרָם מִמִּצְרַיִם
one has to abandon the literal meaning, "went up/rose", and accept the
borrowed meaning, "migrated back". the text does NOT imply that the
ancients perceived egypt as physically lower than canaan. and in gen 23:8
שְׁמָעוּנִי, וּפִגְעוּ-לִי בְּעֶפְרוֹן בֶּן-צֹחַר
"hit/harm" is rejected and "entreat" is adopted for obvious reasons.
when we form a phrase, our words try to approximate a certain image; more
often than not, if our phrasing is not perfect, we rely on the listener's
capacity to reconstruct the same image in spite of its imperfection. but the
listener must cooperate by flexibilizing his definitions of the words. i think
that "(L PNY HRQY(" in gen 1-2 implies a necessarily touchable
literal surface only under a very inflexible reading.