Ancestral Humans Could Speak,
        Anthropologists' Finding Suggests
 

        By JOHN NOBLE WILFORD

        While scientists agree that speech is probably the most important
        behavioral attribute that distinguishes human beings from other animals,
        they have been at a loss to determine when and how that transforming
        evolutionary step occurred.

        They have probed the human brain and compared it with casts of the
        braincase from ancient fossil skulls. They have compared bones and muscle
        attachment points in the throats of humans, apes and ancestral human
        skeletons. Archeologists have examined patterns in early stone tools for
        clues to when humans might have developed the creativity and the
        self-awareness usually associated with communication skills like speech.
 

        All they had been able to agree on is that the earliest unambiguous evidence
        for human speech is found in the cave art and other artifacts, particularly in
        Europe and Africa, that began appearing some 40,000 years ago.

        Now scientists at Duke University have explored a new avenue of fossil
        anatomy and found surprising evidence suggesting that vocal capabilities
        like those of modern humans may have evolved among species of the Homo
        line more than 400,000 years ago.

        By then, their research shows, human ancestors may have had a full modern
        complement of the nerves leading to the muscles of the tongue and so could
        have been capable of forming speech sounds.

        The new findings, moreover, indicate that the Neanderthals, relatives of
        modern humans, could have had the same gift for speech. Their extinction
        about 30,000 years ago has often been attributed in part to speech
        deficiencies, restricting their ability for cultural innovation.

        In a report being published today in The Proceedings of the National
        Academy of Sciences, the Duke anthropologists say that if their
        interpretation involving the tongue nerves is correct, ''then humanlike
        speech capabilities may have evolved much earlier than has been inferred
        from the archeological evidence for the antiquity of symbolic thought.''

        The research was conducted by Dr. Richard F. Kay and Dr. Matt Cartmill
        at the Duke Medical Center in Durham, N.C., with the assistance of a
        former student, Michelle Balow. The results were also described earlier
        this month in Salt Lake City at a meeting of the American Association of
        Physical Anthropology.

        ''This is evidence for the proposition that Neanderthals could talk,'' Dr.
        Cartmill said in a telephone interview on Sunday. ''Did they sound like
        modern humans? I don't know.''

        Anthropologists familiar with the research said the findings were
        interesting and exciting. Some were reserving judgment, but not Dr. Erik
        Trinkaus, an anthropologist at Washington University in St. Louis, who
        specializes in Neanderthal studies.

        ''I think it's not only a reasonable conclusion,'' he said, ''but one long
        overdue.''

        Dr. Trinkaus said previous research had been based on deficient anatomical
        reconstructions, none of which adequately took into account the
        neurological aspects for controlling the vocal track to allow for speech. As
        for the possibility of speech by archaic Homo sapiens 400,000 years ago,
        even before Neanderthals, he said this was consistent with a significant
        enlargement of brain size in that period, the appearance of a more complex
        tool technology and migrations into colder climates, where life probably
        depended on greater planning that could be related to advances in
        communications skills.

        On the other side, Dr. Philip Lieberman of Brown University, an authority
        on early language, has argued that the Neanderthal throat would not have
        been well suited for the production of the vowels a, i and u. But Dr.
        Trinkaus contended that a species would not have needed to produce all
        those sounds in order to have speech and language.

        Even the discovery in Israel a decade ago of a Neanderthal skeleton with a
        large hyoid bone, which is in the throat and associated with speech, had not
        settled the issue of Neanderthal speech. Scientists had said there was still
        insufficient fossil evidence to enable an understanding of how the large
        hyoid bone might have influenced the production of vocalizations.

        Dr. Cartmill himself cautioned that the new evidence for earlier human
        speech ''is suggestive, but, in the present state of our knowledge, it is not
        proof.''

        Other scientists noted that other, independent evolutionary developments,
        including a lengthened larynx, enlarged prefrontal brain lobes and some
        reconfigurations of the brain, would have been critical to the emergence of
        speech. The size of the brain of Neanderthals was well within the range of
        that of modern humans.

        The Duke scientists directed their research at the hypoglossal canal, a hole
        at the bottom of the skull in the back, where the spinal cord connects to the
        brain. Through the canal run nerve fibers from the brain to the muscles of
        the tongue.

        It occurred to the scientists that the size of the hypoglossal canal might
        serve as an index of the vocal abilities of modern and early humans. The
        wider the canal, they assumed, the more nerve fibers there could be to
        control the tongue muscles. And the more nerves, they suggested, the finer
        control the species could have over its tongue for the purpose of making
        speech sounds.

        The researchers compared measurements of hypoglossal canals of modern
        humans, apes and several human ancestor fossils, and concluded that the
        canals of modern humans are almost twice as large as those of modern apes
        -- the chimpanzee and the gorilla -- which are incapable of speech. They
        also found that the canal size of australopithecines, earlier human relatives
        that died out about one million years ago, did not differ much from that of
        chimpanzees.

        The results, the scientists reported, ''suggest minimum and maximum dates
        for the appearance of the modern human pattern of tongue motor innervation
        and speech abilities.''

        To narrow the range, the scientists examined skeletons of Neanderthals and
        also of species of the Homo genus that lived as much as 400,000 years ago.
        These included Kabwe specimens from Africa and Swanscombe fossils
        from Europe. Their hypoglossal canals fell within the range of those of
        modern Homo sapiens.

        ''By the time we get to the Kabwe, about 400,000 years ago, you get a canal
        that's a modern size,'' Dr. Cartmill said. ''And that's true of all later Homo
        species, including Neanderthal.''