Believe it or not, the answer to your question is something that historians themselves have been arguing over for quite literally decades. There's even a very famous (and contentious) book by Edward H. Carr with this very title - What is History?, published in 1961, and a lot more has been written on the topic since.

For my part, I would begin by saying that it is important to be aware that there is an enormous and significant distinction to be made between 'the past' and 'history', even though the two are often used in day to day conversation as though they're synonyms. The past is something that actually happened; a physical occurrence or reality that was once immediately tangible and is now lost through the passage of time. History is our understanding of and about the past. To put it in simple terms: my going to the shop to buy milk this morning is something that happened in the past. But a paper that tells you why I went to the shop, how I went about buying stuff there and what the significance of my trip was to the rest of my day would be a history of what happened. A work of history essentially seeks to answer some or all the essential questions we all have about any past time or event that interests: who, when, where, what, why and how?

History, in essence, is an attempt to reconstruct the past and usually to give it some kind of meaning and context - and though you certainly can have history that is purely narrative and seeks only to tell you what happened, who it involved and in what sequence events unfolded, the very process of writing history (deciding what and who to mention and in what fashion) ends up imbuing it with some kind of meaning even if it is not spelled out in the text. Anyone can make history about anything and anyone. Indeed, many of us in some ways are engaged in the creation and recording of history all the time, in the sense that many of us inherit and continue - even if only in a very limited fashion - a tradition of oral history about our own families. And certainly, we are all living histories; through our culture and our community, we each inherit not only the legacies of the past that came before us but also the legacy of how we have understood that past. Your sense of national identity, for example, is very much shaped and influenced by how the history of your nation and the people that came before are articulated to you. Few people would probably have a sense of national pride if they were only ever told histories about the awful things people from their nation had done.

By implication though, history has the capacity to be - and most would argue, is intrinsically - flawed. Until someone invents a time machine, we have no means of getting back to the past or of making it a tangible reality again; instead, we must try to reconstruct the past from what we have left of it, from the echoes that carry into our own time. There are a variety of ways in which you can do that - you can, for example, write a history book that simply brings together the work of other historians and synthesises them together. Ultimately though, even if you write that kind of book, at some point in the chain your findings will come from some kind of trace of the past. At the most basic level, that trace of the past can come in the form of memory; if you were writing a paper on your late great grandmother's life for example, you could do so by turning to the memories of the people who knew her and lived alongside her. More likely though, if you wanted to write a comprehensive book, you would look for physical traces of the past relating to her life: letters she had written and received, payslips from her various employers if she had any, birth and marriage certificates, the clothes she had worn, photographs or paintings of how she looked, objects she had collected, the house she lived in, and so on and so forth. You would want to study and analyse these things to see what they tell you about her. You can build a history from these things.

But if history is just an understanding of the past, something you piece together from fragments of a lost time or person, then surely you can also be wrong about the past. Indeed, when a history book is seen as substandard or lacking in sophistication, we often refer to it as an example of bad history. Rarely do we deny that such a book is an attempt at writing history per se in the way a pharmacist might dismiss, for example, homoeopathy as 'pseudo-science' (i.e., not in fact science at all). Bad history is still a work of history in the sense that it is an attempt to present an understanding of past people and events, but we see the product of that attempt as being too flawed to be useful to us. How, then, do we distinguish 'good history' from 'bad history'? This, I would suggest, is where the discipline of History separates from the practice of history in general.

History as a discipline seeks to establish principles and guidelines that lead us to understandings of the past that are rooted as strongly as possible in the traces of the past we have available to us. Someone involved in the discipline of History uses these traces, which we call primary sources, as evidence for constructing their understanding. They use them in accordance to the historical method - a set of tools and techniques that let us critically assess the evidence we have and construct from it an argument for a particular understanding of the past. This method is essential to writing good history. For example, without it, we would not be able to write about an event if we had two completely contradictory sources describing the same event; the historical method gives us the tools to critically pick apart those sources and try to determine which (if either) record is more useful to us in trying to understand that event. History as a discipline also seeks to understand the broader context in which events and people live; it tries to understand them somewhat as they would have understood themselves, and to understand the significance of the events in which they participated for wider histories. To that end, it requires some measure of distance from what it studies. Consider, for example, the Wikipedia article on the history of Jamaica. From the 1990s onwards, it really isn't much more than a list of a few events here and there (nothing significant happened between 2002 and 2007 apparently); simply put, not enough time has unfolded for us to fully understand the significance of recent events, or even to understand what events are significant and worth talking about most.

This rigid requirement to make arguments rooted in evidence when constructing an understanding of the past, and to offer explanations of the people and events of the past with appreciation for their proper context and their significance, distinguishes the discipline of History from the wider practice of history in general. Whilst anyone can write history, writing good history is something that generally requires training and experience - though one certainly doesn't need to be qualified in an academic sense to write good history and one certainly doesn't need to be a professional academic, that training will usually be provided through at least an undergraduate History programme. Like most things in life, it's a skill-set that needs to be practised and honed. In general, if you want to be a professional historian in an academic context, part of your training will be grappling not only with the issue of what is strong history, but how you write it; certainly here in the UK, any History MA programme will include a module dealing with historical theory, and you will need to pass that module to qualify for your MA and a decent grade to stand a chance at a PhD placement.

None of this is to say that good works of history are above criticism or flawless; rather, a good work of history should hold up to scrutiny if you were to retrace the author's process of engaging with the evidence they have collected in their research. Bad history might miss obvious evidence, or make convoluted arguments that simply do not fit what the evidence says, or fail to engage with alternative interpretations of that evidence. Often, bad history makes use of the historical method in only a limited fashion, failing to practice proper source criticism, or uses it in reverse: it makes an argument and then seeks out evidence for it.