So are we headed for a digital dark age?

 Yes and no.

My little blogpost yesterday about the slow death of the postal service brought to mind some of the problems brought about by the demise of physical documents.

While we can still read a 150 year old postcard we may struggle to read a 10 year old Microsoft chat message. 

For official records this not a great problem.

At a governmental level there have been a range of initiatives to avoid the loss of access to contents due to obselesence of both hardware and software.

And having worked in both digital archiving and digital preservation I know it can be done. But I also know it's not necessarily cheap.

In digital preservation there are two and a half problems - old hardware used to store the data, old and undocumented data formats, and once you've retrieved it, how to stop it happening again.

Data can be stored on a variety of physical media. Physical media has its own problems with degradation especially if it has been stored in a less than optimal manner, but on the whole most media from the last twenty or thirty years can be read if you have access to the correct device to read it.

Of course, access to the correct media is the key - and not all media is the same - take the humble 3.5" disk - could be written in the common DOS style 1.44MB format, the less common early Macintosh variable speed 800k format, or some strange format used by early Apricot computers (Long gone now, but once the flagship of the UK domestic PC industry and widely adopted by government and large hospitals at the time).

And sometimes the media is just weird - such as the strange 3" disks used by Amstrad PCW machines that were wildly popular in the UK in the late eighties and early nineties.

And of course the format that the data is written in can be incomprehensible to modern software, although both AbiWord and LibreOffice are quite good at reading a range of legacy word processor formats.

If you belong to reasonably well funded body, such as a university or government sponsored archive service this basically resolves to a set of technical problems, which when solved, allows you to extract the contents of text documents into a useful format. 

Mail can be trickier, but actually the original Unix mbox format was simply a lot of concatenated message bodies, it's usually possible to snip the messages apart with a little programming.

What however does the amateur or family historian do?

They don't have the expertise, or even access to suitable hardware to read all the notes that Aunty Ethel put together on her old Amstrad, or even her old Mac Classic using Claris Works.

Unless she printed them out and put them in a binder, they're essentially lost, so while we might be able to read Great Uncle Jack's letters, we can't actually read Ethel's notes and information gleaned from talking to family members.

And that's a problem.

Ethel may no longer be with us. If Ethel had done a reasonable amount of work talking to people who actually knew Jack, and perhaps interviewing them semi formally, we might have a valuable resource, which can't be repeated as the people who knew Jack have also passed on.

While the image of the amateur family historian might be of some bumbly old person in a battered cardigan (and that can be uncomfortably close to the truth in some cases) a lot of them do good work and are educated people who had professional jobs and who turned to family history in retirement as a way of maintaining their skills.

So a digital dark age?

For official records, probably not, always of course barring a major catastrophe. For small scale amateur history I think we are as technological drift means the paper trail of letters comes to a halt sometime after 1990 when people started regularly having computers at home ...

The lingering death of the postal service ...

 If, in the future, we ever want a date to mark when the postal service began a terminal decline, today, the 15th of April 2024 would probably do as well as any other.

Today is the day that Australia Post finally admitted that letter volumes were unlikely to ever recover, and reduced street deliveries to three days a week.

To be fair, it’s been a long time coming.

Hardly anyone sends personal letters any more - and most bills are emailed. Postcards, both picture postcards and as a way of sending a simple note have more or less ceased to exist.

No more Aerogrammes or letter cards, and a future generation will not experience the anticipation of waiting for a letter from a lover overseas. All gone.

One could of course wallow in nostalgia, but basically the letter service is gone, and unlikely to ever come back. These days the only things I mail are official documents that for whatever reason cannot be scanned and emailed, or on a couple of occasions I've sent a letter to deliberately circumvent a useless virtual agent on a company’s website where they still published a street address but not an email address or simple plain online contact form with no useless ‘helpful’ AI built in.

But I’m ranting.

The demise of the letter service does however provide a serious problem for archival research. Scanned letters and letter books can be easily read, and people did tend to keep letters, either as keepsakes from family members or for official purposes as proof that something happened or was agreed.

1879 Postcard - still perfectly readable

Email messages, WhatsApp messages and the rest less so. We’ve seen the Covid-19 response enquiry in Scotland grind to an inconclusive halt over missing WhatsApp messages, and I know of one major insurance case where the tapes supposedly holding crucial archived email correspondence proved to be unreadable.

And while people have raised concerns, we’ve never really come to any conclusions. But one thing is certain, while today we can read Great Uncle Jack’s letters home from his time in the International Brigades during the Spanish civil war, we won’t be able to read any of today’s messages home.

Having dabbled in family history, I can see that’s a problem, it’s not just the loss of colour and background to flesh out an individual, often, as in the case of our hypothetical Great Uncle Jack, it might be the only real proof of where he was and what happened to him ...

The Athenaeum ...

 A couple of weeks ago I tooted that I was starting back with the National Trust documenting the contents of Lake View House in Chiltern

Well, that's still happening, but due to a totally unforeseen event, the start of the project has had to be delayed until May, by which time it'll be working in thermals and fingerless gloves if the house is anything like as cold as Dow's pharmacy over winter.

In the meantime I've got another gig at Stanley Athenaeum, which comes under the Mechanics Institute of Victoria, documenting stuff that hasn't been previously documented - for example I'm currently working on the records of an anti logging protest group from the early Nineteen nineties just before the internet became a thing - so we've got faxes instead of emails and voluminous printouts of meeting papers rather than piles of word documents.

It's not all papers though, the Athenaeum has a remarkable book collection dating back to the 1860s, including an 1861 edition of Darwin's Origin of the Species, and other more practical publications such as Victorian guides to chicken husbandry, and neatly showcasing both the Victorian desire for self improvement, but also the need for people to learn new skills in a strange land ...

Mechanics Institutes are themselves quite interesting. In Australia Mechanics Institutes were often a more middle class thing and as well as self improvement provided a forum for people to meet and discuss ideas, which is why some were described as Athenaeums, such as in Chiltern and Yackandandah, or as a School of the Arts as in Wahgunyah, to give some local examples.

In some places in time they became public libraries with the buildings repurposed - Stanley is relatively unique in that, while it did serve as a library for the community the collection was never broken up or absorbed into a larger collection.

AI and family history

 I've dabbled in family history.

I wouldn't ever style myself as a genealogist, but during the pandemic when you couldn't go anywhere or do anything as  both a distraction and as a way of keeping my research skills up to speed I dabbled.

I'm sure I wasn't the only one to do this, and like a lot of people I signed up to one of the online family history behemoths.

Now, the one I signed up to, MyHeritage, had a lot of quite useful link building tools to automate searches and build family trees, and this is probably a situation where AI might be helpful in flagging spurious leads and resolving inconsistencies.

I'd bailed from big online family history companies before AI really became a thing, so I've no direct experience of the tools currently available, but at the time I bailed they were touting ways to colorise old photographs and also to assemble images of family groups by combining various group photographs.

Hopefully the exif data on the resulting images would make it plain that it was a synthetic image, but one can see the potential for mischief - 'oh look, there's great aunt Vera with Trotsky' - and the like.

Especially confusing if indeed great aunt Vera might actually have worked in Trotsky's office - think of the rabbit holes that could lead to.

Then this morning I came across the following toot:

So people are actually generating artificial images and using them as profile pictures for uncle Cuthbert.

At one level it's harmless if there's no identifiable photograph of Cuthbert, on another one it's something that makes me uncomfortable.

Some people in my family tree were undoubtedly black and some were mixed race.  And while I havn't pieced together all the evidence to say one way or the other, I may have Maori cousins.

The use of these AI tools is dangerous as it would allow people to airbrush history.

One of the reasons I have mixed race cousins is that a distant relation of mine owned slaves in Jamaica.

Not a pleasant thing to admit to, but there's no denying history.

And for that reason, I think there needs to be an understanding that manipulated images and generated images need to be clearly watermarked as such - yes of course sticking images together for a Christmas card is fine as a bit of fun, but we've got to be clear that it's only a bit of computer generated fun ...