Echoes of Compassion: A Look into the Speeches of Charles Dickens
This blog post has been contributed by Ellie-Louise Medd, a student at the University of Leeds who undertook an internship focused on using digital methods to explore the speeches of Dickens. Below, she discusses her findings.
Introduction
The primary focus of this chain of investigation was four particular divisions of Charles Dickens’s speeches: those made for educational causes, hospitals, literary funds, and those spoken at Mechanics’ Institutes. Each of these can be found across Dickens’s life and they often share similarities, especially in his use of rhetoric, though the speeches at the Mechanics’ Institutes sometimes act as contrasting examples due to the context.
It is interesting in itself that Dickens speaks quite differently at the Mechanics’ Institutes. Often, his speeches are celebrations of the arts, and/or promote the improvement of social conditions, primarily for children. There are fewer speeches from the Mechanics’ Institutes, and they tend to be speeches that credit the intelligence of the men that attend. The three main principles of rhetoric – logos (appeal to logic), pathos (appeal to emotion) and ethos (credibility) – can be identified through Dickens’s specific word choice, and explored in new ways when we consider the sentiment of his speeches using digital tools. Each speech makes use of lexis and structures that engage his audience, and show him to be an effective speaker. His varied speeches, covering a multitude of topics, show how adaptable he was as both a speaker and activist, especially where children and social reform were concerned.
In an interview for our exhibition, Michael Slater comments, ‘I think Dickens knew his speeches would be very widely reported and quite influential, so I think he would have [... taken] them very seriously. He was obviously a very gifted [...] speaker, partly because he was a very good actor I suppose, and by all accounts his speeches were very funny, very moving. He was very eloquent, very effective, and, of course, he was bombarded with requests to give speeches for this charity, or that charity, or so on’. My analysis aims to show how effective Dickens was and what made him so effective, looking at his use of certain words and creation of sentiment.
I chose to focus on four words that give scope for broad exploration: ‘child/children’, ‘poor’, ‘little’ and ‘creature(s)’. In the process of transcribing the speeches, which was where our process began, I was drawn to Dickens’s ability to evoke empathy and demonstrate compassion for suffering children, particularly in the busy cities of England. As someone who has studied an array of Dickens’s novels over the course of my academic journey, themes and devices were employed by Dickens in his speeches that directly correlate with some of the depictions of children, particularly vulnerable children, in his novels.
My focus was motivated by the regular mentions of children, whether in speeches fundraising for educational institutes or otherwise. Dickens often used anecdotes to represent the experiences of young children, and often referred to the mortality of these children, engaging the pathos of listeners who should care and want to contribute to improving the lives of young people. He also frequently used first-person pronouns, which contribute to the ethos of his speeches: he positions himself as a credible and knowledgeable speaker that his audience can trust. Dickens’s brief personal experience with the cruel world of factory work, and his wide knowledge of the dire state of cities in the wake of industrialisation, is a thread through his literary works, and also underpins some of the speeches. He often reiterated his position on these topics: he was not a politician but a writer, and his skill as a writer is not separate from his skill as a speaker. Catherine Waters and Peter Merchant’s book Dickens and the Imagined Child presents a nuanced understanding of the figure of ‘the child’ across Dickens’s work. They write, ‘Dickens's multifaceted response to childhood no doubt owes much to the diversity of opinion circulating about the topic in his own day’ (5). They also introduce the idea of the ‘reinvention of childhood, towards the end of the eighteenth century, as an ideal state’ (5), which is evident in Dickens’s representations too.
Method
I used AntConc as a starting point for my research, creating individual corpora for each speech topic. I imported and created a corpus for each topic, and began running some terms, and firstly looking at Key Words in Context (KWIC). Using the collocates and n-gram features, I began to formulate a path for my research, and imported the files into R Studio to begin to visualise my findings. It was during this process that I identified and created my own stop word list, as some of the hits in the frequency analysis would show articles like ‘a’, and ‘the’, along with some context-specific words like ‘ladies and gentlemen’, and ‘Prosperity to...’ – phrases Dickens often used in his speeches to introduce or conclude them. The process of creating a stop word list was an evolutionary one: for example, ‘upon’ frequently appeared and needed to be added to the stop words list, as it was a common linguistic feature of the time, and provided nothing in terms of context analysis.
After making the final visualisations of the data in R Studio, including mapping sentiment using the NRC emotional lexicon, I began to make comparative analyses between the datasets, identifying common trends and picking out what was particularly important, or significantly different. For some frequency data, the highest frequency was only 3 (in terms of how many times it appeared in the specific corpus; while this might seem low, a recurrence of a phrase three times in a small body of speeches can still be significant. The NRC lexicon is based on modern text data, which undoubtedly skews some of the sentiment analysis and cannot be properly representative of Victorian attitudes; however, it yielded interesting results and may be worth pursuing in more detail in future.
I also began making comparisons with Dickens’s novels, as I felt the speeches and novels shared a lot of common themes, and seeing these in context allowed for some interesting comparisons. For this I used CLiC.
It must be noted that we are reliant on transcriptions of Dickens’s speeches from different sources of varying quality: Dickens did not speak from notes, and only in some cases corrected transcriptions made of his speeches ahead of publication. The collection we have is drawn from different newspaper reports, diaries, biographies and cannot represent a perfect recreation of what Dickens said. As such, taking a more distant approach and considering keywords and sentiment is a fruitful way into the big ideas and themes of Dickens’s speeches, which remain consistent across different transcriptions, rather than too tight a focus on phrases. At the end of this blog post I explore Dickens’s use of pronouns, but this is an area where transcriptions can differ quite substantially: a ‘your’ can be mis-transcribed as ‘our’ very easily in different newspaper reports. Nonetheless, my aim is to show the possibilities for more sustained analysis of Dickens’s speeches, and what can be done with our newly transcribed collection.
Sentiment Analysis and Most Frequent Words
Both sentiment analysis and word frequency analysis inform each other, and can be compared to understand not only what rhetoric techniques are used but also to understand the reason why certain words are more frequent than others.
Hospital Speeches
Figure 1: Sentiment Analysis in Hospital Speeches
Figure 2: Top 20 Most Frequent Words in Hospital Speeches
Figure 1 shows sentiment analysis of the hospital speeches. It shows that the most common sentiment was ‘positive’, though followed by ‘negative’, and ‘fear’. But, to understand this, I wanted to look at the top 20 most frequent words that appeared in this corpus. Figure 2 displays this information and, notably, there is a repeated motif of fatality with ‘poor’, ‘sick’, as well as more obvious references to ‘hospital’, ‘health’, ‘medical’, and institutions. In looking at the frequent words, the positive sentiment could be informed by words like ‘good’, and ‘better’, and ‘years’ also poses the potential for a positive discussion: mentions of years ago, and now change, or even mentioning change to come in years ahead.
‘Trust’ was a sentiment that appeared in all the speeches, which initially appeared to be an interesting result; however, even though Dickens uses words like ‘honest’ and ‘confident’ to address his listener and promote his reliability, I think this repeated result may have been a result of his repetition of the phrase ‘Prosperity to...’ and other similar formal linguistic features of nineteenth-century speech-making. It is for this reason that I will not be individually analysing why this appears on all the sentiment analyses. I can only suggest that the most organic contributions to ‘trust’ as a sentiment could be appeals to the ethos and logos of the listeners, almost as a form of flattery for support.
Figure 3: Top 20 Context Words Surrounding ‘little’
‘Fear’ as a sentiment is prevalent. I believe Dickens was using children to appeal to the audience’s fear and also to create pathos. On AntConc, most of the hits for ‘children’ were in the 1858 speech for the Hospital for Sick Children (unsurprisingly). ‘Little’ frequently accompanied ‘children’, or was used as an adjective pertaining to them. In the Hospital for Sick Children speech, there were 20 hits alone for ‘little’. This led me to use R Studio to look at the context words for ‘little’. I did not have time in this internship to repeat this process for every single speech, or every single word, but I believe it was entirely relevant for the investigation of ‘little’ in the speeches.
As Figure 3 shows, ‘poor’ was a frequent context word, as well as ‘bed’, ‘body’, ‘child’, ‘sick’ and ‘wan’, all of which create this image of weak and ill children in need of the support of the listeners. This is an evident appeal to the feelings and emotions of the crowds, yet these speeches are not the only instances of the sick or ill children motif in Dickens. A very brief search for ‘little’ in CLiC revealed that ‘little’ is frequently found with ‘boy’, ‘girl’, ‘thing’, and ‘coffin’ (which is also used in this same speech). These are emotive examples of Dickens appealing to the pathos of his listeners, and utilising the feeble, young child as a device as well as drawing on real-life examples of the types of individuals on whose behalf he was speaking.
The context words of ‘poor’ also revealed a very similar patter (see Figure 4), where ‘poor’ is often attributed to ‘lady’, ‘child’ and ‘creatures’, which is mirrored in the novels Dickens often used ‘creature’ in association with a young, sometimes attractive woman, such as Rose Maylie in Oliver Twist. Using AntConc to compare this use of ‘creature(s)’, it showed within his speeches a frequent collocate of ‘creatures’ was ‘suffering’, evoking sympathy and compassion. ‘Creature’ is a loaded word choice, pejorative in nature, and creates something like a wounded animal depiction of the children (and others, as not all of the hospital speeches consisted of discussion of children).
Figure 4: Top 20 Context Words Surrounding ‘poor’
Educational and Literary Speeches
Figure 5: Sentiment Analysis of Education Speeches
Moving to look at speeches focused on education and literary concerns, it is clear to see a shift in the particular sentiments that take precedent. In Figures 5 and 6, anticipation scores much higher and positive still rules as the highest sentiment. ‘Fear’ is much less present in the literary speeches; there is some presence of fear still worth noting in the education speeches, but this could easily be another consequence of a child-centred concern in that corpora. Again, this can further be developed by looking at the Top 20 Most Frequent Words in each of the corpora.
Figure 6: Sentiment Analysis of Literary Speeches
Figures 7 and 8’s visualisation of word frequency provides an interesting insight, particularly the presence of ‘little’ in both. ‘Children’ also appears in the education speeches as one of the top 20 most frequent words, with ‘young’ also present. One of the education speeches took place in 1869, close to the passing of the Education Act of 1870 which established compulsory schooling for children aged 5-13. This could coincide with the frequent use of ‘hope’, suggesting improvement for coming years.
Figure 7: Top 20 Most Frequent Words in Education Speeches
Figure 8: Top 20 Most Frequent Words in Literary Speeches
Looking at the KWIC in context of ‘poor’, it is associated with ‘wretched’, ‘writer’, and ‘inventor’ in the literary speeches, suggesting that Dickens uses ‘poor’ to engage with pathos. The positive and joy sentiments of the literary speeches can be understood to be a result of Dickens’s organic interest in the literary speeches, as a writer himself. Those speeches certainly contain a livelier literary interest, replete with Shakespeare references and appeals to ‘Friend’, ‘friends’ and ‘committee’, placing Dickens in a more personal relation to the cause. ‘Literature’ takes the top spot in the 20 most frequent words.
It would be worth investigating this use of ‘literature’ further. One notable use was in his speech at the Royal Academy Banquet on 30 April 1870, his last recorded speech: Dickens uses ‘literature’ in the context of ‘the brotherhood of Literature’, which is a powerful depiction of where he is speaking from. ‘Brotherhood of Literature’ is impactful; it has connotations of almost religious bonds, something consistent with Dickens’s lifelong commitment to the literary world. It also suggests a shared, common interest in literature, showing its high value to Dickens. He is creating a collective with the listeners and attendees. My next section will explore collectives, by looking at pronoun use and the bigrams or ‘I’, ‘you’ and ‘we’.
‘I’, ‘You’, and ‘We’ Bigrams
For this particular question, I used the n-gram function on AntConc and then used R Studio to visualise the results. Bigram searches can vary on R Studio, so by using AntConc, I had more control over the data. In visualising the data, I combined the results of ‘I’, ‘you’ and ‘we’. This way, I was handling the top 20 most frequent words in total, instead of having to compare data that was separated into ‘I’, ‘you’ and ‘we’, which would have been inefficient.
Figure 9 to 12 demonstrate these top 20 bigrams of the pronouns in each speech. Among all the data sets I dealt with, ‘I have’ is the most frequent bigram, and ‘I am’ is the second most frequent. The use of the ‘I’ pronoun can typically be related to the ethos and credibility of a speaker, drawing to the listeners attention – their position. The use of the ‘I’ pronoun is more a cause for comparison, as it is expected that Dickens would speak in first person. Just taking a look at the KWIC for the education speeches on AntConc shows the different contexts in which ‘I have’ is actually used. In the education speeches, ‘I have’ is seen with ‘I have been’, and ‘I have no doubt’ and ‘I have never seen’, which are active statements. He is both presently and previously involved with the situations to which he refers. This is visible through the use of ‘I am’ and ‘I do’, also – he is, naturally, speaking in the present tense.
Figure 9: Top 20 Combined Bigrams (Education)
Although ‘I’ is obviously extremely common, I was intrigued to see how Dickens presented collectives, and how he directed the speech, using ‘we’, and ‘you’. In the context of the ‘we’, ‘we have’ appears to be the most common in the literary and educational speeches. In the education speeches, the collocates of interest for ‘we have’, are ‘building’, and ‘company’. He uses ‘we have’ when referring to the strengths of the institution or company, making it a group concern and creating a collective interest and investment in the topic, with himself a part of it. In the literary speeches, he does a similar thing with ‘we have’, as he uses it to highlight what they have done together: ‘we have set forth’, ‘we have met here’, ‘we have done’, and ‘we have also felt’. This is a creative way of fostering shared emotions and presenting a united front, gathered over sentiments as well as actions.
Figure 10: Top 20 Combined Bigrams (Literary)
In the Mechanics’ Institutes speeches and hospital speeches, ‘we are’ is the most common bigram for the ‘we’ pronoun, and it is again used to create a collective identity: ‘we are interested, ‘we are assembled’, ‘we are met to promote’. In the hospital speeches, he uses ‘we are’ to highlight shared knowledge: e.g. ‘we are not without our experience’; ‘we are perfectly acquainted’. One of the more interesting usages of ‘we are’ appears in the speech of the 9 February 1858, in which he uses ‘we are’ to describe his audience as ‘in a childlike state altogether’. This combines his interest in children and his attempts to create a collective identity. He goes on, in this same sentence, to discuss ‘representing an infant institution, and for the sake of children, acknowledge that we have an interest in them’. Here, there is also the bigram of ‘we have’ present.
Figure 11: Top 20 Combined Bigrams (Mechanics’ Institutes)
Figure 12: Top 20 Combined Bigrams (Hospital)
The use of the ‘you’ pronoun across the speeches varies slightly more, and is more complicated to analyse. A prominent usage is ‘you may’, which Dickens can be often seen using as a strategy to instruct and inform, e.g. ‘you may see/recollect/learn/bear testimony’. Across Figures 9 to 12, he appears to use ‘you have’ frequently. In the literary speeches, for example, ‘you have’ is used gratefully, either to express his own personal gratitude for the funds and societies – ‘you have done me the great honour’, or ‘you have honoured us tonight’ – or to credit the particular audience he is speaking to – ‘you have heard so eloquently’, ‘thank you for the tribute you have paid to it’, or ‘you have made that discovery’. ‘You’ is a very personal and direct address to his audience which works in tandem with the use of ‘we’ in his speeches: he is closing the gap between himself and those present to listen to him.
Conclusion
There is still much that can be explored. With Dickens’s speeches being so understudied and never before available digitally in one place, the scope for analysis is endless. What I have concluded from my brief analysis of the data is that Dickens’s had a profound and deep-rooted concern for children, and the social conditions in which they lived, learned and died. His speeches align and create parallels with his literary work, for which CLiC is an invaluable comparative resource. The topics I focused on demonstrate Dickens’s sophisticated use of rhetoric to make appeals to his audience, to promote change, and to advocate for improvement. His speeches present his genuine concern for the future of education and hospitals, and also recognised the importance of Mechanics’ Institutes and literary institutions – something you can tell he genuinely loved. So little scholarship exists on the speeches, and I hope that more will be said about this fascinating body of work that encompasses Dickens’s entire professional career.