"Each tRNA is specific for one amino acid, and it's what recognizes the codon."
Yes, but WHY does a tRNA code for an amino acid? What is there about it that says that this group of nucleotides "means" this amino acid?
Are you asking if there's any reason, beyond accidental, for the association of specific codes with particular amino acids? There are some speculative ideas floating around about this, but little experimental evidence.
The molecular structure of each different type of tRNA is such that it bonds preferentially to a particular amino acid.
A few semi-relevant papers:
Genetic Code Origin: Are the Pathways of Type Glu-tRNAGln \rightarrow Gln-tRNAGln Molecular Fossils or Not?Abstract: A logical-evolutionary analysis is conducted to clarify whether or not pathways of type Glu-tRNAGln -> Gln-tRNAGln are molecular fossils of the mechanism that gave rise to the evolutionary organization of the genetic code. The result of this analysis is that these pathways are most likely a manifestation of this mechanism. This provides strong evidence in favor of the coevolution theory of genetic code origin, as this theory is based on the amino acid biosynthetic transformation taking place on tRNA-like molecules which imprinted the genetic code structuring. Comments on the different interpretations of these pathways found in the literature are also provided.Phylogeny from Function: The Origin of tRNA Is in Replication, not TranslationExcerpt: Here we propose a phylogeny for the origin of tRNA based on the ubiquity and conservation of tRNA-like structures in the replication of contemporary genomes, and we discuss the evidence in contemporary molecules that leads to and supports this phylogeny. The unique aspect of this phylogeny is that it places the origin of tRNA in replication, before the advent of templated protein synthesis. This implies that tRNAs arose before the other components of the translation apparatus, that aminoacyl-tRNA synthetases arose next, and that both tRNA and aminoacyl-tRNA synthetases predated the anticodon and mRNA.[...]
Transitional Genomes as Clues to Early Replication Strategies. If the original role of tRNA-like structures was in replication, as suggested by the single-stranded bacteriophage and plant virus genomes, one might expect to find additional examples of contemporary genomes in which tRNA plays that same role. When we first proposed the genomic tag hypothesis (Weiner and Maizels, 1987), only one other example of tRNA involvement in replication was known: In modern retroviruses, tRNAs function as primers for initiation of cDNA synthesis by the retroviral reverse transcriptase. Over the past few years, additional novel replication strategies have been described that employ tRNA-like structures. These appear to link replication of single-stranded RNA viruses with retroviral replication and with the synthesis of modern chromosomal telomeres. In each of these instances, a genomic RNA replicates via a DNA intermediate. We call these "transitional genomes," because they can be viewed as reenacting the transition from an RNA world to the contemporary DNA world.