The Harappan language is the unknown language or languages of the Bronze Age (2nd millennium BC) Harappan civilization (Indus Valley civilization, or IVC). The Harappan script has long defied attempts to read it, and therefore the language remains unknown. The language being unattested in any readable contemporary source, hypotheses regarding its nature are reduced to purported loanwords and substratum influence, notably the substratum in Vedic Sanskrit and a few terms recorded in Sumerian cuneiform (such as Meluhha), in conjunction with analyses of the undeciphered Indus script. There are a handful of possible loanwords from the language of the Indus Valley civilization. Sumerian Meluhha may be derived from a native term for the Indus Valley civilization, also reflected in Sanskrit mleccha meaning non-Vedic or native, and Witzel (2000) further suggests that Sumerian GIŠšimmar (a type of tree) may be cognate to Rigvedic śimbala and śalmali (also names of trees). There are a number of hypotheses as to the nature of this unknown language: One hypothesis places it within or near the Dravidian languages, perhaps identical with Proto-Dravidian itself. Proposed by Henry Heras in the 1950s, the hypothesis has gained some plausibility and is endorsed by Kamil Zvelebil, Asko Parpola and Iravatham Mahadevan. A 2021 research paper published in Nature clarifies that Proto-Dravidian was spoken in Indus Valley based on the ultraconserved Dravidian tooth-word and genetics. a "lost phylum", i.e. a language with no living continuants (or perhaps a last living reflex in the moribund Nihali language). In this case, the only trace left by the language of the Indus Valley civilization would be historical substratum influence, in particular the substratum in Vedic Sanskrit. Hypotheses that have gained less mainstream academic acceptance include: an Indo-European language, close or identical to Proto-Indo-Iranian: suggested by archaeologist Shikaripura Ranganatha Rao.