Comment on "Linguistic Features of Noncoding DNA Sequences"
In a recent Physical Review Letter, Mantegna et. al., report that certain statistical signatures of natural language can be found in non-coding DNA sequences. In this comment we show that random noise with power-law correlation similar to $1/f$ noise, exhibits the same "linguistic" signature as those found in non-coding DNA. We conclude that these signatures cannot distinguish languages from noise.
adap-org↗