View of memory
If in writing code, one thinks in terms of instructions and data which can be understood by a processor (eg: bytes in the main memory, registers, hardware IO signals), one is working with a low level language.
High level languages allow the programmer to think of memory locations and data more abstractly/ declaratively by starting with assumptions that there exist memory locations with certain properties which will be bound to variables used by the programmer by the compiler with the OS’s help.
Mixed level languages allow a combination of both these paradigms.
Portability
Programs written in high level language are more portable across computers with different architectures and abilities due to the existence of compilers or interpreters or virtual machines implemented on various platforms which suitably translate a high level program to something understandable to the processor.
Programming discipline
High level languages often try to enable/ enforce some programming discipline (eg: by enforcing OOP paradigm, functional programming, variable scoping etc..), provide more concise commands for frequent operations, perhaps using ’syntactic sugar’; though internally they may be forced to use programming paradigms different from the programming interface provided to the programmer.
Type system
Implementing abstract data-types
While programming, rather than think of data in terms of bytes, one can consider and manipulate data (collection of data) at a more abstract level. One can define different generic operations and views on a set of bytes depending on the type of data stored therein.
Examples of data-types may be Integers, Floating point numbers or even arbitrary user-defined classes. A very basic datatype offered by high-level languages is the Array/ sequential storage.
Compile time vs run-time
Aka Static vs dynamic.
In static type checking, data types are associated with variables, and these are checked at compile time to ensure that assignments do not violate this binding. So, the binding of a type to a variable happens at compile time, rather than run time. This enables us to catch errors and prove more properties about a program’s correctness before it is compiled and run. It also enables safer refactoring of code (changing type to a subtype etc..).
In case of dynamic typing, types are associated with values rather than variables. THis allows the programmer more flexibility in his use of variables.
Strong vs weak
Can the type of data pointed to by a variable symbol change within its scope?
Scala has weak typing and static typing. Python has weak typing and dynamic typing.