WEBVTT 00:00:00.240 --> 00:00:06.339 Hello. In the Crown of Aragon 639 years ago, 00:00:06.879 --> 00:00:13.400 it was February 9th, 1387. My name is Jonathan 00:00:13.400 --> 00:00:16.679 Seyfried. I'm a PhD candidate in history at the 00:00:16.679 --> 00:00:20.480 University of New Mexico. And this is the Historian's 00:00:20.480 --> 00:00:24.980 Notebook, a podcast about how history gets made. 00:00:25.660 --> 00:00:29.890 Season one is titled Molt Cara Companyona. We're 00:00:29.890 --> 00:00:32.729 looking at a document from each day of the first 00:00:32.729 --> 00:00:37.049 year of the reign of King Joan I of Aragon and 00:00:37.049 --> 00:00:41.030 Queen Violant de Bar. Let's begin today's document. 00:01:05.260 --> 00:01:08.060 At the start of the episode today, I want to 00:01:08.060 --> 00:01:13.159 give a shout out to my dad. It's his birthday 00:01:13.159 --> 00:01:18.900 today. Happy birthday, dad. All right. So in 00:01:18.900 --> 00:01:22.000 addition to it being my dad's birthday, this 00:01:22.000 --> 00:01:27.579 is also the date of a little letter in the secret 00:01:27.579 --> 00:01:34.000 seal register of Violant de Bar. This is one 00:01:34.000 --> 00:01:37.260 of the shorter letters that we'll be working 00:01:37.260 --> 00:01:41.959 on this podcast season. I think it's probably 00:01:41.959 --> 00:01:45.980 the shortest one thus far at about this is at 00:01:45.980 --> 00:01:48.680 four lines so I don't think we've had a document 00:01:48.680 --> 00:01:53.579 this short before and when I saw it I thought 00:01:53.579 --> 00:01:56.620 to myself oh this this won't be too difficult 00:01:56.620 --> 00:02:01.719 it's very short oh no there was a challenge with 00:02:01.719 --> 00:02:06.120 this letter because i cannot for the life of 00:02:06.120 --> 00:02:11.520 me figure out who it's addressed to and i looked 00:02:11.520 --> 00:02:14.639 at the folio before and i looked at the letters 00:02:14.639 --> 00:02:18.139 following they do not seem to be related they're 00:02:18.139 --> 00:02:24.759 on different days and this one is the only February 00:02:24.759 --> 00:02:31.060 9th letter around for Violant. So even the dating 00:02:31.060 --> 00:02:34.659 of it is weird because the scribe has chosen 00:02:34.659 --> 00:02:40.180 to put down the letter nine as Roman numeral 00:02:40.180 --> 00:02:49.500 five followed by four i's. So V -I -I -I -I. 00:02:50.090 --> 00:02:52.930 Which is not usually what I see. Usually when 00:02:52.930 --> 00:02:56.090 I see the number 9 rendered in these documents, 00:02:56.389 --> 00:03:02.849 it's rendered as IX for the Roman numeral 9. 00:03:03.099 --> 00:03:07.699 So there's some mystery here in this letter, 00:03:07.699 --> 00:03:13.900 and I turned to good old chat GPT in order to 00:03:13.900 --> 00:03:16.759 get some help with it. And it turns out that 00:03:16.759 --> 00:03:21.139 this is a really good letter for practicing one's 00:03:21.139 --> 00:03:25.639 conjugations in medieval Catalan, because there's 00:03:25.639 --> 00:03:33.139 a lot of uses of the verb venir to come and so 00:03:33.139 --> 00:03:38.240 we're seeing it these conditional phrases like 00:03:38.240 --> 00:03:46.039 if you would come or please come to the court 00:03:46.039 --> 00:03:49.340 so we've got well let me see i'll just like list 00:03:49.340 --> 00:04:01.009 them off here we've got I think those are the 00:04:01.009 --> 00:04:05.169 three that appear here. So the way that this 00:04:05.169 --> 00:04:11.569 letter starts is a typical initial phrase, Entes 00:04:11.569 --> 00:04:16.730 havem, which means we have understood. So Violant 00:04:16.730 --> 00:04:21.490 is starting out this letter by saying, hey, we've 00:04:21.490 --> 00:04:25.670 gotten some information and here's what the information 00:04:25.670 --> 00:04:32.180 is. Then... We have a person's name a little 00:04:32.180 --> 00:04:38.199 bit further down the line Miguel de Noval's But 00:04:38.199 --> 00:04:41.199 what comes right before that is the phrase per 00:04:41.199 --> 00:04:49.339 relacio den and in the modern normalization of 00:04:49.339 --> 00:04:53.480 Den there would be an apostrophe so D apostrophe 00:04:53.480 --> 00:05:01.160 N so that would expand out to day and which is 00:05:01.160 --> 00:05:07.139 of mister so what the first few words of this 00:05:07.139 --> 00:05:12.660 letter are saying is we have understood by the 00:05:12.660 --> 00:05:20.100 report of mr miguel de noval's and then following 00:05:20.100 --> 00:05:28.980 the name is que which would translate to that And 00:05:28.980 --> 00:05:35.680 then we have vos vendriez, so that you would 00:05:35.680 --> 00:05:42.699 come. So Miguel de Noval has reported to Violant 00:05:42.699 --> 00:05:48.519 that the addressee of this letter would come 00:05:48.519 --> 00:05:55.220 to Barcelona. And then the rest of the letter 00:05:55.220 --> 00:06:00.990 is just kind of ornamenting that by saying phrases 00:06:00.990 --> 00:06:05.930 like oh if you if you knew that you that i wanted 00:06:05.930 --> 00:06:09.670 you to come then you would definitely come so 00:06:09.670 --> 00:06:15.730 what i'm doing is i'm letting you know nos certificam 00:06:15.730 --> 00:06:20.589 i am informing you that i would be pleased if 00:06:20.589 --> 00:06:26.879 you came to our presence here if you came to 00:06:26.879 --> 00:06:31.980 the the court and and that's it so basically 00:06:31.980 --> 00:06:36.240 what this boils down to is i heard from miguel 00:06:36.240 --> 00:06:39.459 de noval's that if you knew that i wanted you 00:06:39.459 --> 00:06:42.480 here you would come here and so i'm telling you 00:06:42.480 --> 00:06:49.259 i want you here but the question is who and We 00:06:49.259 --> 00:06:54.720 don't know. The probata does not give us any 00:06:54.720 --> 00:07:00.019 clues. It simply is your typical probata. Dona 00:07:00.019 --> 00:07:05.500 Regina mandauit mihi is the first line of the 00:07:05.500 --> 00:07:11.480 probata. And then pere de basanta. Pere de Basant 00:07:11.480 --> 00:07:14.220 is the scribe. We see a lot of letters in this 00:07:14.220 --> 00:07:18.180 register that have been copied by pere de basant. 00:07:18.399 --> 00:07:23.959 And then the probata p at the end. So there's 00:07:23.959 --> 00:07:30.120 no clues here. And I honestly am really surprised 00:07:30.120 --> 00:07:35.579 because one of the points of being able to record 00:07:35.579 --> 00:07:38.899 these letters is that you can come back to them 00:07:38.899 --> 00:07:44.279 and access their meaning. And so the idea that 00:07:44.279 --> 00:07:48.350 they would let this one go without putting in 00:07:48.350 --> 00:07:52.529 the addressee in the line below the last line 00:07:52.529 --> 00:07:57.629 as often happens is, to me, hard to understand. 00:07:58.569 --> 00:08:03.709 Maybe they were just in a hurry? I don't know 00:08:03.709 --> 00:08:07.250 what to make of this. It doesn't really seem 00:08:07.250 --> 00:08:11.129 like a situation where there would be some kind 00:08:11.129 --> 00:08:16.529 of secrecy required, so... I don't know. I'm 00:08:16.529 --> 00:08:19.290 just gonna have to hope that at some other point 00:08:19.290 --> 00:08:24.310 I find another mention of Miguel de Novales. 00:08:24.350 --> 00:08:29.810 I did go to chat GPT again to see if Miguel de 00:08:29.810 --> 00:08:33.669 Novales appears in any of the secondary literature. 00:08:34.269 --> 00:08:39.470 There is an article by a scholar named Guillermo 00:08:39.470 --> 00:08:45.669 Lopez Juan. from just 2024 in which Miguel de 00:08:45.669 --> 00:08:51.610 Novales is mentioned as a member of the oligarchy 00:08:51.610 --> 00:08:59.190 in Valencia. So that's something and yeah I'll 00:08:59.190 --> 00:09:05.809 just have to keep on looking around but I think 00:09:05.809 --> 00:09:11.860 what's important here is the fun dancing tone 00:09:11.860 --> 00:09:17.139 of this little note like why is it necessary 00:09:17.139 --> 00:09:22.980 to have all this conditional phrasing to say 00:09:22.980 --> 00:09:29.139 I summon you or who is Violant writing to that 00:09:29.139 --> 00:09:34.889 she feels the need to couch her request for this 00:09:34.889 --> 00:09:39.669 person's presence in this kind of language. It's 00:09:39.669 --> 00:09:44.370 pretty interesting. So I'm really having a sense 00:09:44.370 --> 00:09:49.149 here of building the plane as I'm flying it on 00:09:49.149 --> 00:09:53.450 its initial voyage, because my understanding 00:09:53.450 --> 00:09:57.639 of the nuance of medieval Catalan is really at 00:09:57.639 --> 00:10:01.059 the beginning stages. This summer I'm taking 00:10:01.059 --> 00:10:08.980 an intensive seminar in medieval Catalan, and 00:10:08.980 --> 00:10:12.480 I hope to get from that a little bit more understanding 00:10:12.480 --> 00:10:15.720 of some of the nuances like what I'm seeing in 00:10:15.720 --> 00:10:19.080 this particular letter. And suffice it to say 00:10:19.080 --> 00:10:22.320 that it's always an ongoing learning process. 00:10:22.639 --> 00:10:25.679 You know, this is how history gets made. You 00:10:25.679 --> 00:10:28.960 try to make progress on your understanding of 00:10:28.960 --> 00:10:32.080 what happened in the past as you're overcoming 00:10:32.080 --> 00:10:35.899 language barriers or other kinds of barriers 00:10:35.899 --> 00:10:40.659 and understanding that will arise when you're 00:10:40.659 --> 00:10:43.700 trying to imagine what was going on in such a 00:10:43.700 --> 00:10:47.379 long time ago or even not so long ago. You can 00:10:47.379 --> 00:10:52.210 have the assumptions and the perceptions that 00:10:52.210 --> 00:10:57.389 ground you in the present day interrupting your 00:10:57.389 --> 00:11:03.330 understanding of what happened in the past. a 00:11:03.330 --> 00:11:07.169 little letter like this can bring that out in 00:11:07.169 --> 00:11:10.850 some surprising ways and so what historians do 00:11:10.850 --> 00:11:14.070 is they do their best and then later historians 00:11:14.070 --> 00:11:17.750 will come and look over again a particular letter 00:11:17.750 --> 00:11:21.129 or a particular interpretation and they'll say 00:11:21.129 --> 00:11:25.350 hey actually I can refine this and that's how 00:11:25.350 --> 00:11:28.009 real history gets made and it's actually like 00:11:28.009 --> 00:11:31.399 a lot of fields of knowledge like can imagine 00:11:31.399 --> 00:11:34.159 what I'm talking about in terms of scientific 00:11:34.159 --> 00:11:38.080 experiments where someone does a scientific experiment, 00:11:38.480 --> 00:11:41.620 they reach a certain conclusion, and then someone 00:11:41.620 --> 00:11:44.779 else comes back later and challenges that with 00:11:44.779 --> 00:11:49.340 maybe a larger sample size in their study, so 00:11:49.340 --> 00:11:54.340 on and so forth. So knowledge building as a collaborative 00:11:54.340 --> 00:12:00.799 process is key to how history gets made, and 00:12:00.799 --> 00:12:04.539 it's key to understanding the value of expertise 00:12:04.539 --> 00:12:09.440 overall. And that leads me to discussing a little 00:12:09.440 --> 00:12:11.639 bit about AI. I thought this would be a good 00:12:11.639 --> 00:12:16.700 episode for me to get into some of the guidelines 00:12:16.700 --> 00:12:22.200 that I have in mind when I'm using AI, because 00:12:22.200 --> 00:12:25.940 it's incredibly helpful, but you've got to be 00:12:25.940 --> 00:12:29.230 very cautious. And so I'm going to explain a 00:12:29.230 --> 00:12:33.090 little bit about my cautionary notes. First of 00:12:33.090 --> 00:12:36.570 all, what you have been noticing if you're looking 00:12:36.570 --> 00:12:40.690 at a lot of the AI chat transcripts is that I 00:12:40.690 --> 00:12:46.649 have not done anything to limit the sycophantic 00:12:46.649 --> 00:12:52.600 tone that AI is generally programmed to give 00:12:52.600 --> 00:12:57.460 its users. So my go -to chat bot is ChatGPT, 00:12:58.080 --> 00:13:03.659 and famously it is known for this kind of positive 00:13:03.659 --> 00:13:07.539 reinforcement saying things like, ooh, you did 00:13:07.539 --> 00:13:12.240 a good job on that, or having these weird emotional 00:13:12.240 --> 00:13:16.460 responses like saying, ooh, this letter is so 00:13:16.460 --> 00:13:22.080 very Joan, which is bizarre. Like I can't even 00:13:22.080 --> 00:13:25.299 figure out why it wants to do that. But it's 00:13:25.299 --> 00:13:30.840 part of this tone programming that the folks 00:13:30.840 --> 00:13:35.799 at open AI think is going to make people more 00:13:35.799 --> 00:13:40.039 excited to use chat GPT. I find it annoying and 00:13:40.039 --> 00:13:44.799 in some cases very like grating like it's irritating 00:13:44.799 --> 00:13:49.850 but You know, I keep it on and the reason I have 00:13:49.850 --> 00:13:52.710 not done it you can change your settings to eliminate 00:13:52.710 --> 00:13:55.990 that But I want to keep it on because I want 00:13:55.990 --> 00:14:01.649 to be reminded that I'm dealing with this entity 00:14:01.649 --> 00:14:05.509 that can sound authoritative and I don't want 00:14:05.509 --> 00:14:08.950 to fall into the trap of believing everything 00:14:08.950 --> 00:14:14.590 I hear so if I keep that level of positive feedback 00:14:14.590 --> 00:14:20.440 sycophancy turned on, then I'm routinely reminded 00:14:20.440 --> 00:14:25.039 that what I'm being presented with might sound 00:14:25.039 --> 00:14:28.440 comprehensive when it's really not, and it might 00:14:28.440 --> 00:14:32.440 sound authoritative when it is not justified 00:14:32.440 --> 00:14:38.259 to be an authority on a particular topic. Now, 00:14:39.080 --> 00:14:43.889 that said, I have found that in this project 00:14:43.889 --> 00:14:49.889 in particular I have been saved by chat GPT in 00:14:49.889 --> 00:14:52.909 terms of unlocking the meaning of the document 00:14:52.909 --> 00:14:57.230 or being alerted to aspects of the document that 00:14:57.230 --> 00:14:59.429 are really interesting to me that I wouldn't 00:14:59.429 --> 00:15:02.149 have picked up on if I was just doing this on 00:15:02.149 --> 00:15:06.659 my own without the. assistance of the AI chatbot, 00:15:06.919 --> 00:15:11.580 including today's letter, my knowledge of medieval 00:15:11.580 --> 00:15:14.279 Catalan grammar, or just, you know, even modern 00:15:14.279 --> 00:15:17.960 Catalan grammar, isn't really good enough for 00:15:17.960 --> 00:15:23.179 me to have really pulled out some of the grammatical 00:15:23.179 --> 00:15:26.279 layers of meaning for this document. So even 00:15:26.279 --> 00:15:30.309 if chat GPT is wrong in certain areas of its 00:15:30.309 --> 00:15:33.649 interpretation, I'm still being alerted to that 00:15:33.649 --> 00:15:35.950 there's something going on here that I could 00:15:35.950 --> 00:15:41.549 look into more deeply. And then the other other 00:15:41.549 --> 00:15:44.429 part of this is that I can see what chat GPT 00:15:44.429 --> 00:15:49.370 is searching. So if I ask it to search for secondary 00:15:49.370 --> 00:15:54.159 literature from scholarly sources, I can see 00:15:54.159 --> 00:15:56.700 the websites that it's looking at and I can see 00:15:56.700 --> 00:16:00.659 that it's going to publishers websites like Brill 00:16:00.659 --> 00:16:04.379 or Cambridge University Press. Publishers that 00:16:04.379 --> 00:16:10.679 I know are extremely high quality in the world 00:16:10.679 --> 00:16:15.299 of academic publishing. So I'm able to monitor 00:16:15.299 --> 00:16:19.740 what ChatGPT is doing in its searches. It absolutely 00:16:19.740 --> 00:16:23.210 could have some blind spots and you know, the 00:16:23.210 --> 00:16:29.149 whole copyright violation issue with the AI chatbots 00:16:29.149 --> 00:16:34.429 and all of the scraping of data is a real ethical 00:16:34.429 --> 00:16:40.169 issue. And I'm not trying to excuse that or gloss 00:16:40.169 --> 00:16:47.610 over it. But there is a cost benefit situation 00:16:47.610 --> 00:16:51.679 that I'm feeling here that I can apply. to the 00:16:51.679 --> 00:16:55.779 ethical stance a little bit, that the gain in 00:16:55.779 --> 00:16:59.580 the practice and the educational endeavor is 00:16:59.580 --> 00:17:03.759 so great that it is worth engaging with the product 00:17:03.759 --> 00:17:07.660 even with its substantial ethical flaws at this 00:17:07.660 --> 00:17:13.339 time. So it's something that I am not dismissive 00:17:13.339 --> 00:17:15.920 of in any way and I keep returning to and thinking 00:17:15.920 --> 00:17:18.960 through and my monitoring to see what happens 00:17:18.960 --> 00:17:23.740 in the industry and how certain products their 00:17:23.740 --> 00:17:27.559 companies are more or less responsive to the 00:17:27.559 --> 00:17:30.079 feedback they're getting about the ethical dimensions 00:17:30.079 --> 00:17:34.940 of what they're doing. So that was my little 00:17:34.940 --> 00:17:40.279 take on chat GPT. and using AI, there's more 00:17:40.279 --> 00:17:42.740 to be said, but I wanted to address it just for 00:17:42.740 --> 00:17:45.299 a little while, especially because this was a 00:17:45.299 --> 00:17:47.619 pretty short letter, and there was a little bit 00:17:47.619 --> 00:17:51.660 of time in today's episode to do that. Thanks 00:17:51.660 --> 00:17:54.480 so much for listening to this episode of the 00:17:54.480 --> 00:17:58.160 Historian's Notebook, Season 1, Molt Cara Companyona. 00:17:58.660 --> 00:18:00.779 If you are leaving with more questions than you 00:18:00.779 --> 00:18:04.720 arrived with, I've done my job. Remember, the 00:18:04.720 --> 00:18:08.909 motto of the Historian's Notebook is dissatisfaction 00:18:08.909 --> 00:18:12.309 guaranteed. Visit the website to see an image 00:18:12.309 --> 00:18:15.210 of today's document. It's the little tiny one 00:18:15.210 --> 00:18:19.130 at the very top of the folio. And also on the 00:18:19.130 --> 00:18:22.089 webpage you'll find additional show notes. And 00:18:22.089 --> 00:18:24.589 listen again tomorrow to hear about the next 00:18:24.589 --> 00:18:28.569 day in the first year of the reign of King Joan 00:18:28.569 --> 00:18:32.630 of Aragon and Queen Violant de Bar. In the 00:18:32.630 --> 00:18:34.410 meantime, take care.