Academic history is a conversation. Sometimes the conversation is good, sometimes it's not, but it does - like other conversations between people - have a dorection as it continues across generations. What that means is that while newer is not always better, it almost always makes the most sense to start with recent, up-tp-date contributions to the conversation rather than interjecting something feom 50 (or 500!) years ago as though the subsequent years of discussion were pointless.

For example: a lot of people like to cite Gibbon's Decline and Fall of the Roman Empire in newspaper editorials. It's an amazing, magisterial work, and it still shapes many of the questions that historians discuss today. But it's also more than 200 years old, and problems arise when you quote it as an alternative to a recent, current work. The most basic is that you ignore the fact that, for 200+ years, people have been reacting to and critiquing Gibbon, and many of these critiques are valid. That's less about teleology (newer is not always better), and more about the fact that 200 years of smart people reacting to something another smart person said invariably leads to new insights and different perspectives. Writing those perspectives off because Gibbon's work was 'great in its own right' (it was) shows a disregard for the kind of critical reactions and further study that has allowed us to refine, respond to, and in many cases reject as unfounded the evocative statements of Gibbon's eloquently written volumes.

More importantly, reading Gibbon as a timeless classic makes it easy to miss how much his version of decline and fall was shaped by contemporary (18th century) anxieties about the future of the British Empire in which he was living. Gibbon's not unique in that - every work of history is written in a time and place, and is shaped by contemporary concerns. Part of whybhistoriography matters is that, ad time goes on, we do a lot of hard intellectual work to question how previous scholars' conclusions were shaped by the evidence, and how much their ideas responded to contemporary concerns; and which of their ideas, nevertheless, remain valuable and instructive today.

That's why the most recent books are usually the first place to start. Not because newer is always better,mbut because the most recent books (if they're written well) will be engaging with the ones that came before them, and showing how new works fits in with older classics. How the conversation has advanced. And it sometimes happens that you'll decide the conversation took a wrong turn (there's an article from 1950 that basically sets up my PhD work, but no one did anything like it until me, about 60 years later), and you find yourself bringing up an old classic (or, more often, a dusty old scholar who was unfairly ignored 100 years ago) and telling people it was important. But when you do this, you're not throwing out the conversation that happened in between - you're not being that jerk who, in a lively conversation, keeps trying to drag people back to a point made an hour before as though everything everyone said since was a waste of time (that's a great way to drag down a conversation). Instead, you're taking something old and showing what it can add something useful to the present community by reading it against what people are talking about right now.

Newer isn't always better, but the conversation has almost always moved on from where we were when old classica were written. And while some good ideas may have been lost from this, a lot of bad ones were too. It's not helpful to try to rewind the discipline; if you want to use old sources, you need to be able to show how they fit into the current state of knowledge, and to distinguish between their useful insights and the things that, now, were clearly only useful for the book's contemporary, original audience.

History is more than just a collection of facts arranged into a narrative. It's factual veracity is always limited by the sources at our disposal and the analytical tools we have to engage with those sources. In pretty much any given field of history, our understanding is better than previous generations. I'm not a historian of Rome or Egypt, but I am certain that a modern history of Cleopatra will be using archaeological and textual sources which were simply not available to a writer of the 19th century. Therefore the modern historian can draw on more primary resources than their predecessor. This even applies to more modern history, as new archival material is always being unearthed or releases, giving us insights on relatively recent history which those writing history in its aftermath were not aware of.

Apart from sources and analytical tools, there is also the slow collaborative efforts of successive historians drawing on their predecessors' work. Any individual historian can only do so much on s topic, eventually someone with a fresh insight will come to add to even the greatest historian's findings.

If the 19th century history of Cleopatra you mentioned is a groundbreaking and instrumental text in that particular area of history, then I can guarantee that it will have been evaluated by the groundbreaking historians who came after it, who would have retained the best scholarship from that text in their work, while discarding the weaker scholarship.

Another thing to bear in mind is that history is not written in a vacuum. It is necessarily influenced by the historian's own historical context, their biases, and their understanding of the world. The biases and analytical methods of the past may be alien to us today, or have been proven insufficient by more recent findings and theoretical frameworks.

So to give an example, in my own field, there's a very important work, Lorimer's Gazette. This multi-volume text from about a century ago was written by a British colonial official, who over ten years compiled a history of the emirates on the Persian Gulf coastline, including geographic and demographic information. This Gazette is a very important record, as it gives a thorough history of diplomatic and military developments over the course of 200 years. However it is also a classic Orientalist text, and was commissioned in order to understand the Gulf Arabs and exert greater control (indeed, the text coincides with a heightened period of British colonial activity in the region).

So the Gazette is quite useful for anyone studying 18th and 19th century Arabia. It is an invaluable record, and can be generally assumed to be accurate (particular for its record of the years following British-Arab treaty relations). However it is also steeped in the context of the Gazette's purpose and the frankly wrong assumptions. For instance, it explains the Wahhabi conquests of the 18th century as being successful in part because they practised a more correct version of their faith. This essentialist portrayal of Islam would never be accepted today. Some British recorders of history describe the Arabs as lazy compared to Persians, while others will state that the Persians are lazy, unlike the wilful and active Arabs.

Modern histories of the Gulf monarchies still rely on Lorimer. But they discard the incorrect and biased information and sift out the relevant facts.

From my own Cold War area research, historiography changes focus a lot over time. The first great debate was from the 1960s to the 1980s over 'what are the origins of the Cold War? and Who is responsible for the Cold War?', split between Orthodox, Revisionist and Post-revisionist positions.

If I was looking towards answering that question I could largely end my journal access in 1995 and still do a good job (though new challenges are emerging from scholars inspired by Ted Hopf). The 'big question' in the field at the moment is more about 'What is the Cold War?' split into a conservative camp who want to remain largely focused on superpowers, the military and diplomacy of the Cold War, and various factions of 'broadeners' (who want to explore how the Cold War rivalries were interpreted and manipulated by the smaller states, and how things like culture, social and intellectual history were impacted), and finally the 'deepeners' who want to place the Cold War into long term trends like imperialism and colonisation and orientalism.

The older historiography can be excellent, it really depends on what you want to answer, and there are certainly excellent possibilities in returning to a'long settled' debate, with a fresh eye.

Another thing is that if newer things are presumed to be better (and in some sense they have the potential to be), then that leaves opportunities open for current academics to make their names out there. Which they obviously have incentive to do. When a new work is published, if it has any impact of course everything before it slightly loses relevance, because their marketshare has gone down.

A 160 year old book is kind of stretching it though. That's like the time when American historians were talking about the great patriotic war and the unifying spirit of America colonizing and civilizing the untamed West. Not really great for study of anything but the perceptions people had at the time.

Follow-up question: Have there been periods in which historiography became worse than it had been previously?