Documentation of archaeological and cultural heritage sites is at the heart of the archaeological process and an important component in cultural heritage research and presentation. It is an essential step without which interpretation and analysis are not possible. It is what makes archaeology and cultural heritage “scientific”. Maybe we are storytellers. If so, the type of story we tell is heavily influenced by our way of collecting and organising our archaeological data. But can we speak about CORRECT documentation or should we talk only about usable and non-usable documentation? The contemporary field is plagued by the involvement of operators each with their own new tools. They propose solutions and suggest methods but are often in blissful ignorance of the past investigations of the item, site or cultural heritage they are working on. New technology, however, has to support our research. Its use still depends on what we want to know next (our research). The best solution is to have an underpinning of basic documentation that allows any new researcher to easily access the core record. Then they can then enrich the documentation with the results of their new method, analysis and ideas. It may be possible to build the ultimate recording system, but the information we feed it is always potentially unreliable. How do we know when our record is good – has integrity? What indicates that it might be bad – lacking integrity? Models are there to be used, not believed. Documentation is always for a certain purpose and depending on that purpose, a set of documentation may be regarded as good or bad, as “fit for purpose”. There will never be absolute “true”, “correct” or “right” documentation. An abstract model of documentation should consist of the attributes we record of the real world traits that we observe. The set of attributes that we choose to record (out of the infinite set of possibilities) are the ones that our current state of knowledge and our research aims (and therefore designs) suggest will be the most useful to our current research aims/agenda. If we want to reuse data beyond the current research project/agenda then we must be very explicit about why, how and what we record. This is the so called “para” data and goes beyond meta data to include the “how” and “why” of data capture.