Voice recognition

This is an experiment.

For a long time I’ve wanted to start blogging again and for various reasons. But the timing isn’t particularly good at the moment because last September I quit my job as an ALT to go freelance as a translator, proofreader, possibly interpreter, and ­– very tentatively – writer. So I’ve been spending a lot of time tapping away at a computer keyboard and in fact, I already have mild RSI as a result. I’ve been adjusting how I use my computer and bought a new, ergonomic computer keyboard, which is a funny shape and supposed to be good for your hands, wrists, and arms. But I started a job the other week that involves a lot of translating in a short space of time and in particular, inputting the times in a video at which people start talking. On the first day of this, I did getting on for 10 hours of translating almost without a break and filled well over 300 cells in an Excel worksheet, every single one of which needed a time of six figures, as in 00 hours 00 minutes 00 seconds. That day in particular put a huge strain on pretty much every part of my body, including my legs and my feet, and the next day it felt like I’d run a marathon.

A few years ago I suffered what in Japanese is called gikkurigoshi (ぎっくり腰/a bad back). The physio told me I was on the way to slipping a disc and thanks to her treatment and advice, I still do back exercises. While I haven’t bought a standing desk yet, I do stand up to work at the computer.

So I was wondering how I could start blogging again without putting undue strain on my already tired typing fingers, and I’ve heard from one or two translators that using voice recognition software is a good way of doing so. It would seem the technology has advanced quite significantly, particularly for English. (If you use voice recognition for Japanese – which is, relatively speaking, a minor language – it’s liable to make more mistakes.) So as I speak, I’m holding my smartphone in one hand and dictating into Microsoft Word. When I’ve finished dictating, I’ll save the document, transfer it to my laptop, edit it, and post it on the blog. Which all sounds fine and dandy, but in practice, this is the second time I’ve tried this. The first time, what happened was that I spent an hour dictating a post about mountain biking, which I will probably post after this one, and once I had transferred it to the laptop, I spent at least two or three hours editing it. That didn’t necessarily involve rewriting, as I wanted to save time and effort, but it did involve deleting extraneous words, correcting the words and phrases that the voice recognition feature in Microsoft Word had misheard and misinterpreted, then tweaking. Forgive me father, for I have tweaked!

Even having edited and deleted and tweaked, the resulting post is unsatisfying to say the least. It does make sense and it also makes me realise the little habits, tics, words, phrases, or grammatical constructions that I use too often when I speak. But more than that, there’s just something…weird about it.

One of the very nice people who I deal with at Camphor Press – which is going to publish my cycling book, Charinko: 2,500 Kilometres Around Japan by Bicycle, in May of this year – said that he writes their email newsletter using voice recognition software and that when he does so, he uses bullet points. So he doesn’t just start recording and go – which is what I’m doing now – he thinks about what he wants to write, makes a few notes and based on those notes, starts recording. I didn’t ask him how long it takes to edit the newsletter from the voice recognition file, but I wonder how much time he’s actually saving.

The book was originally due to be published in time for the 2020 Olympics and for various reasons – not necessarily procrastination on my part or the part of the publishers – it’s been delayed. There’s no particular hurry because the book itself is about a tour that I went on in 2005 – i.e. 17 years ago. I want to use the blog as a method for promoting the book and have been advised to start an email newsletter myself. I haven’t actually looked into how or whether WordPress can handle email newsletters, but I will in due course.

So I suppose what I’m saying is, please forgive me if this and subsequent blog posts over the next few months and years don’t have quite the same writing brilliance to them (said he, ironically) as before. For example, the last post that I wrote was in the summer of 2021. I got a very good response, particularly when I shared it on Facebook, but it didn’t really count as a blog post because I’d spent many, many hours writing, rewriting, editing, and yes, tweaking it.

One of the original purposes of blogging on the internet was spontaneity. In particular, I always think of my favourite Japanese blog, Igirisu Dokuzetsu Nikki. The Japanese woman who writes it lives in England with her husband and two children. She writes very eloquently about cultural differences and the inconveniences and contradictions of British culture and British life, as well as about her slightly crazy in-laws. She doesn’t have repetitive strain injury, but she has some kind of physical problem which means that she can only spend 15 minutes on a computer each day. So every evening for the past 15 or 20 years, she’s been writing blog posts for 15 minutes at a time. Despite that limitation – perhaps because of that limitation – her blog is brilliant: funny, perceptive, and well written, to the extent that her Ameblo blog ranking is very high and a few years ago she was contacted by a company that wanted to publish a compilation of her best posts. I always wonder in cases like this whether she will be found out, because a lot of the things that she writes about are potentially libellous. Although she writes in Japanese and anonymously, she names the town where she lives and you wouldn’t have to be a detective to find out who she was. If you were one of her friends or neighbours and you decided to go online, find her blog and feed it into Google Translate or DeepL, you might come across something that you, your friends and your neighbours didn’t want to read.

But I digress. What I want to do with my new voice recognition strategy is to be more spontaneous, to blog more regularly, and to not be quite so fussy about the writing quality. I want to post more often, generate page views, and generate interest in the book. So I apologise if you think that you’re being used in some kind of marketing ruse. I’d also like to say that if anybody is reading this post, I very much appreciate the fact that you are and despite my almost complete lack of motivation for promoting Muzuhashi and for interacting with other bloggers, I’ve been blessed with a disproportionately large amount of positive feedback. I hope that you will continue to read, even though the posts herein may from now on make you feel slightly uneasy, in the sense that they may not read like they were written by me, the reason being that they will have been dictated, then edited as hastily as possible with aching wrists and fingers, and posted without the same rigorous quality control that until now I like to think I’ve been using. So please bear with me. I do hope that in future I can start using bullet points and dictating more concise posts that make more sense.

I also hope that I can give you some insights into the background to the book, as my current idea is to produce a kind of director’s commentary, so that people who have read it and even people who haven’t can get an insight into the tour itself, my state of mind when I embarked on it, what actually happened as opposed to what I wrote about, and the process of writing the book and getting it published. I would also like to post bonus material and passages that didn’t make it to the final edit, and/or research that I didn’t use.

My intention is to produce a less rambling version of the usual DVD director’s commentary, and rather than a continuous commentary on the book, to give you posts relating to particular chapters and particular incidents that will hopefully act as a companion to – and as something that enhances your enjoyment of – the book. But we shall see. Like I say, this is an experiment and whether I continue with it or how I adapt it to suit my needs and the needs of my readers, I don’t yet know. If nothing else, please look forward to more blog content.

Most bloggers have at least one post that says something like, “Sorry that I haven’t been blogging lately. I’ve been very busy. I keep on meaning to and I don’t get round to it, but I promise I’m going to blog more often.” This is my version of that post and whether I can keep my promise to blog more often, I don’t yet know. Whether I will end up writing or dictating another post almost identical to this in six months or a year’s time saying, “Sorry I haven’t been blogging lately. I’ve been very busy,” I also don’t yet know.

So this is me dictating into my smartphone using voice recognition as an experiment. We shall see how that experiment goes over the coming months and years.