"It's like a time machine," said Altschuler, an associate professor at the University of Medicine and Dentistry of New Jersey.
Some traditional classics scholars are dubious of this statistical approach, described in the journal BioEssays by Altschuler and Mark Pagel, an evolutionary biologist at the University of Reading in England.
Yet the date Pagel and Altschuler came up with is pretty close to what many historical linguists already believe. Traditional efforts to date the Homeric epics have relied on such clues as textual references to customs and objects known to be from the eighth century B.C., said Sheila Murnaghan, a professor of Greek at the University of Pennsylvania.
Altschuler, 44, practices rehab medicine at University Hospital in Newark, N.J., helping patients with back pain, knee arthritis - and, yes, even pain in the Achilles tendon. He reminds his students to spell it with a capital A, telling them it is the proper name of the Greek hero.
When not at work, Altschuler maintains a passion for ancient languages, a throwback in an era when many Americans associate the name Homer with the dopey dad on The Simpsons.
Altschuler studied Latin at New York's Stuyvesant High School, followed by Greek and Hittite at Harvard.
He got the idea for dating Homer after reading a 2007 paper by Pagel about the rates at which words were replaced. For example, the word dog has largely replaced the Old English word hund, though we still have the similar, or cognate, word hound.
Altschuler contacted Pagel, and the two agreed to collaborate. The physician gathered the data, assembling and verifying the lists of 173 common words from modern Greek, Hittite, and Homer. Pagel and his Reading colleagues did the statistical analysis.
As might be expected, languages separated by longer periods of time were less similar.
Of the 173 words, just 23 remained cognate (like hound-hund) between Hittite and modern Greek. Hittite and Homeric Greek were closer, with 33 cognates.
The greatest similarity was seen between Homeric and modern Greek. About half of the words, 87 out of 173, were cognates.
Hittite, Homeric Greek, and modern Greek all are descended from an ancestral Indo-European tongue spoken thousands of years ago, though Hittite is on a different branch of the tree.
Don Ringe, a linguistics professor at Penn who was not involved with the work, took issue with several aspects of the analysis. Among them: The Greek of Homer is not one language, he said, but an "artificial literary dialect" containing aspects of at least two languages. Moreover, some of Homer's words were hundreds of years out of date when the poems were put down on paper, Ringe said.
Altschuler said that even if that were true, he doubted it would affect the analysis, because the researchers studied common words that are slow to change.
There also is some debate as to whether the epics were set to paper at the time Homer composed them, or if they were written down long after his death. And there are those who question whether the epics are truly the work of one person, or of several. Altschuler scoffs at that notion, arguing that the work hangs together as a coherent, lyrical whole.
"It's too good," he said.
In addition to applying statistics to language, Altschuler has used it to study the music of Bach. In 1998, he wrote a book called Bachanalia: The Essential Listener's Guide to Bach's "Well-Tempered Clavier."
His medical research includes an effort to find survivors of the 1918 flu, whose immune systems still bore antibodies to that lethal virus. He also has been to Haiti to treat people who lost limbs in the 2010 earthquake, using a technique called mirror therapy to rid them of "phantom pain" in arms or legs that no longer exist.
He and Pagel hope others will apply their linguistic method to Beowulf and other ancient works. People already have written with other suggestions, such as the Mahabharata, an Indian epic in Sanskrit.
Altschuler's heart, however, lies with Homer.
"We want to know. We want to understand," he said of the poet. "I feel like we can almost touch him. I don't know if we're there."
Contact Tom Avril at 215-854-2430 or email@example.com.