A genetic language family tree has as 'leaf' nodes those languages with no descendants, either because the language is living or because it became extinct without leaving any descendants.  Any of the tree's other (rootward/ancestor) nodes/languages was in use at the time that its speakers split into isolated sub groups such that the language of the subgroups diverged into different child languages. 

A rootward/ancestor node/language is known to have existed
  1. through a written historical record or
  2. by inference.  In this case the language is a 'proto' language.
In the following, we discuss case 2.

The family tree (German Stammbaum) genetic model , uses a comparative method to establish [as much as possible of] a proto-language from which known descendant languages and/or daughter proto languages evolved.   The appelation/prefix proto is typically used to call attention to the fact that the language is a constructed prototype of what actually existed.  The prefix proto is used as part of the name of the constructed language itself.

Though substantial fragments of a proto-dictionary and a proto-grammar exist, they came from theorizing linguists rather than from records of speakers.  So proto languages are inferential/speculative/presumptive/theoretical rather than factual/documented/observed.

The associated language family tree showing 'genetic' relationships most or all of which were derived using this comparative method has many successes, though there are complaints and other approaches.

Successes:

The following family trees are widely accepted.  The ancestor languages are well-established and generally not in dispute.

Complaints and other Approaches:

Two of several other approaches are:
Learn  more.



References:

The little book by R.M.W. Dixon titled "The rise and fall of languages" is recommended to learn more.


Our Language Tree:

We have used an available tool, limited to the display of one tree, to


Our language tree has root-most nodes as follows. 
(Every other, leafward, node corresponds to a particular language.)

Root
Each leaf of this tree is a living language (there are ~7,300) or an extinct language with no descendants.   Certain nodes, including this root, are not languages.


  World Proximity Region
A proximity region is a linguistic region (based primarily on geography) for placement of Absolute Isolate languages.
For example:  Paupan -- language absolute isolates of the western Pacific
For example:  Trans-New Guinea -- Paupan absolute isolates  spoken in New Guinea and neighboring islands
The World Proximity Region is the root of a (sub)tree providing successively finer linguistic regions, one of which is most suitable for any Absloute Isolate.
 
 
  Proto-Human-languages
    Every genetic cap-parent is a child of this node. There is no evidence that any of these shared a proto-language.
 
 
  Unclassified
   Children of this node are language leaves lacking any better placement in the language tree.  List of nodes that belong here.

 
               

A linguistic region is a weak grouping of languages based simply on a shared living space.  The living space is physical/geographic, such as a continent.  Proximity (of speakers of the child node languages) implies that their ancestors may have also been physically proximate and therfore may have shared a common genetic proto-language.  A linguistic region should be distinguised from a linguistic area/sprachbund/convergence area/diffusion area which corresponds to a group of languages that have become similar in some way -- because of geographical proximity.

A genetic parent is a node from which child node languages evolved over time.

A genetic cap-parent is a node which, despite linguistic study, has no plausible genetic parent from which it and other proto-languages descended.  Every 'genetic cap-parent' is a child of the Proto-Human languages node.

A relative isolate is a known language known to derive from a distant proto-language but not known to derive from any of the child proto-languages.  For instance, Albanian, Armenian, and Greek are commonly called 'Indo-European isolates'.

An absolute isolate is an intensely studied natural languages with no relation to any proto-language.  Ref  Commonly cited examples include Basque, Ainu, Burushaski, and Korean.


Other Language Trees:


The Ethnologue provides a plausible complete tree similar to ours (and from which we've drawn extensively).