An Empirical Study on the Nature of the Verbal Responses of Humanist Chaplains

Abstract

There has been a limited amount of empirical research conducted in the past on how chaplains, and humanist chaplains in particular, actually interact with their clients during conversations as a part of spiritual care and counselling. The aim of the current study was to gain insight into the extent to which the verbal responses of humanist chaplains corresponded to Rogers’ nondirective approach during conversations with clients. Rogers’ approach has been commonly embraced since the beginning of the professionalization of humanist chaplains in the Netherlands. The study focused on humanist chaplains working at a general hospital in the Netherlands. Ten humanist chaplains took part in the study by audio recording their conversations with clients. The audio recordings were transcribed and analysed, and the verbal responses of humanist chaplains were compared to Rogers’ approach. Subsequently, the verbal responses were analysed via conversation analysis, which also provided insight into how the humanist chaplains actually conversed with clients. Most of the verbal responses (73%) were consistent with Rogers’ nondirective approach, though the ways in which some of the verbal responses were expressed were different; they were more compassionate and comforting. The remaining 27% of the verbal responses were directive and did not correspond to Rogers’ approach. The study shows that, compared to Rogers’ nondirective approach, the approach of the humanist chaplains was more direct and comforting.

1. Introduction

Sometimes humanist chaplains have the feeling that they are left empty-handed. They feel that they have too little to offer to people who are confronted with suffering and the loss of perspective (Mooren 2013b). Chaplains of other denominations recognize this uncertainty in their own practices. They note a lack of expertise within the profession as to how to conduct a conversation (Stone 2001). This could partly be due to a lack of described methods on how to have a conversation (Stone 2001). Despite the fact that various ways of conducting a conversation are available within the profession (see, among others: Jorna 2008; Mooren 2013a; and Smit 2017), the call for acquiring basic conversation skills is more pressing than before1. Mooren (2013a) is of the opinion that, to be effective, chaplains need transferable experience and knowledge within the profession. In other words, a proven basic method is indispensable for chaplains. Oddly enough, very little empirical research has been conducted in the profession regarding the manner of conversation.

There is, nonetheless, modest but growing interest in this type of research within the profession (Fitchett 2017). It often focuses on publications that describe what chaplains should do (Handzo et al. 2008). Other studies are usually based on case studies. In the Netherlands, for example, a Dutch Case Studies Project in Chaplaincy was conducted that describes what chaplains do and what the effects are on clients (Walton and Körver 2018). Research was also conducted on the actions that hospital chaplains take during conversations with clients (see e.g., Handzo et al. 2008).In the study by Handzo et al., hospital chaplains recorded what they did after their conversations with clients (on the basis of 19 prescribed categories). Although these descriptive studies are of value for professionalization purposes, very little empirical research has been conducted on how chaplains actually proceed in their conversations, interventions, and consultations. One example is the study by Handzo et al. (2008), which shows that the chaplains reported that most of their conversations with clients (72%) were based on empathic listening. No studies were conducted on whether they actually put this into practice.

Around the middle of the 20th century, chaplains adopted Rogers’ nondirective approach (see, among others: Rogers 1951; Stone 2001; O’Connor 2003; Snodgrass 2007; and Carey and Cohen 2015). Until the 1950s, people mainly talked about pastoral care with a prominent place assigned to the Proclamation of the Word (Mooren 2008; Carey and Cohen 2015). The chaplain was mainly there to turn the human being before him towards the doctrine of faith (Mooren 2008). In the 1950s, people spoke more and more about “pastoral counselling” (O’Connor 2003). This term was put forward in 1949 and symbolized an increasing embrace of Rogers’ approach, an approach derived from humanistic psychology (Elliot et al. 2004).

As a result of the rise of pastoral counselling, there was a shift in emphasis from the Word Proclamation towards a people orientation: “Today, the dogmatic orientation of the past is opposed to pastoral care as a hermeneutical process [...] as a search for meaning, for the understanding of the other” (Mooren 2008, p. 47). This shift was partly based on a zeitgeist circulating around the Second World War: a shared tendency to turn away from that which was moralistic in nature (Snodgrass 2007; Carey and Cohen 2015). In 1963, the Foundation Humanistic Educational Institution2 was established in the Netherlands, in which chaplains were educated on a humanistic basis. Humanist chaplains assist individuals who are confronted with existential questions and conditions, regardless of their religious background or philosophy of life. They are looking for strength in the clients themselves and provide proximity and attention. Humanist chaplains are inspired by humanistic values, such as individuality, autonomy, equality, and open perception. According to the Dutch Humanist Association, approximately 240 registered humanist chaplains are employed in the Netherlands in various fields, such as hospitals, nursing homes, psychiatry, prisons, and the army. The Association for Spiritual Care in the Netherlands (VGVZ) estimates that 20–25% of all Dutch chaplains work on a humanistic basis. From the beginning of the professionalization of humanist chaplains in the Netherlands, Rogers’ approach was embraced and regarded as the appropriate way of conducting conversation (Mooren 2008). This approach fit like a second skin, especially since humanism is averse to proclamation (Mooren 2013b). Despite the increasing focus on Rogers’ approach around the mid-20th century, very little empirical research was conducted on conversation techniques in chaplaincy (Carrigan 1966). Rogers himself believed that empirical research was inevitable: “Only by careful study of the recorded interview–preferably with both sound recording and transcribed typescript available–it is possible to determine what purpose or purposes are actually being implemented in the interview. ‘Am I actually doing what I think I am doing?’ [...] Only an objective analysis of words and voice can adequately determine the real purpose the therapist is pursuing (Rogers 1951, p. 25)”.

The aim of the current study was to gain insight into the extent to which the verbal responses of humanist chaplains corresponded to Rogers’ nondirective approach via audio recordings of conversations between humanist chaplains and clients at a general hospital. Verbal responses were defined as the audible responses of the humanist chaplains. We were interested in the spoken words, and also in how humanist chaplains were heard and how they expressed themselves. The audio recordings of conversations between the humanist chaplains and clients were the basis of our research since we were not present during the conversations. The main reason not to be there was that the presence of the researcher could hinder the conversation (Carrigan 1966). Nonverbal responses were, therefore, not taken into account. The research question was: What is the nature of the humanist chaplains’ verbal responses during counselling conversations with clients?

2. On Rogers’ Nondirective Approach

The nondirective approach is intended to help the client “to become truly human, because the client must become who he is” (Rogers [1961] 2006, p. 132). Inevitably, this means that a man could be different from what he really is; he could, as it were, not be his true self. In other words, Rogers believed that there could be a discrepancy between who someone is and who someone seems to be. Rogers called this kind of discrepancy “incongruence” (Rogers 1951, [1961] 2006; Van Kalmthout 2005). Rogers’ aim was to close this discrepancy (Rogers [1961] 2006, p. 132). According to Depestele (2008), the therapy is not focused on the relationship between the client and the counsellor, but on the relationship between the client and himself. In other words, the client must be open to who he really is.

Rogers’ approach is nondirective (see, among others: Rogers 1951, [1961] 2006; Brodley 1990, 1997, 1999; and Cheung 2014). This nondirective approach relates to two fundamental assumptions: (1) The clinician must trust the client; and (2) It helps if the client feels understood by the clinician (Brodley 1990). According to Brodley (1990, 1997), Rogers’ nondirective approach is unique, as other approaches are directive, which means that the counsellor is seen as an expert. In Rogers’ view, the expert is not the counsellor, but the client (Rogers [1961] 2006). The counsellor is also described as a ‘’midwife’’ because he is there to help the client give birth to what he was already carrying (Van Kalmthout 2005, p. 135). For Rogers, the nondirective approach is linked to a certain basic attitude.

Rogers’ approach includes three characteristics that he felt are necessary and sufficient to help the client (Rogers 1957): (1) Unconditional positive regard; (2) Empathy; and (3) Congruence (Rogers 1957; Brodley 1990, 1999). This primary attitude is accompanied by verbal responses that are nondirective and that match the client’s frame of reference. According to Brodley (1999), nondirective responses can be divided into “empathic following responses, to verbally indicate that the client can be traced, and “empathic understanding responses”, to request clarification from the client. The verbal responses that represent the frame of reference of the humanist chaplain are characterized as “directive”. Here, too, a distinction is drawn between “allowed directive responses” and “unallowed directive responses”. According to Brodley (1999), we can only speak of an allowed directive response in the following situations:

• During arrangements and while creating the terms of the therapy;
• When addressing questions and requests;
• The moment the counsellor suspects that the client wants to make a comment or ask a question;
• If the counsellor makes empathic observations;
• When making corrections for the loss of acceptance or empathy or incongruence;
• The moment the counsellor obtains an insight or idea that may be of value to the client;
• If the therapist has impulsive, spontaneous, emotional reactions, without the counsellor having a specific intention in advance;
• For spontaneous responses that are based on what the counsellor already knew about the client;
• During spontaneous agreements between the client and counsellor;
• For impulsive, spontaneous, evaluative responses of the therapist. These verbal responses are the result of the counsellor’s proximity to the client and should not interfere with the client;
• If the counsellor tries to put into words an unspoken feeling of the client. However, if this occurs too often, we should no longer speak of a nondirective approach;
• While correcting misunderstandings.

Brodley (1999) refers to allowed directive responses because they are allowed within Rogers’ approach. If a verbal response from a humanist chaplain is “unallowed directive” in nature, it does not correspond to Rogers’ nondirective method. The four types of verbal responses, as distinguished by Brodley (1999), are shown in Figure 1.

3. Method

Because of a lack of empirical research into the ways of conducting conversations within humanist chaplaincy, a qualitative study was chosen (Boeije 2014). Since this research focuses on the nature of the verbal responses of humanist chaplains, the research was discursive in nature. A content analytical approach was utilized, specifically including conversation analysis, whereby the ways in which humanist chaplains conversed with clients were investigated (Have 2007; Boeije 2014). For example, an empathic understanding response could be used in different ways, and conversational analysis helped to explore the differences. Finally, the verbal responses were analysed in an “open” inductive manner, leading to additional themes, represented in a codebook.

3.1. Participants

The population of this study consists of humanist chaplains in the Netherlands who work at a general hospital. To participate in the study, the humanist chaplains had to be willing to make at least one audio recording of a conversation with a client. To prevent the chance that very few verbal responses of the humanist chaplains would be heard, an audio recording needed to have a minimum duration of 15 min. The audio recording also had to be intelligible, and the nature of the conversation had to be in no way linked to certain conditions. For this reason, each participant was allowed to decide what kind of conversation he or she recorded.

A total of twenty-three humanist chaplains were approached for the study, and ten of them participated. The others were asked to indicate why they did not participate in the study, and the following points were mentioned:

• Six said they did not participate because of a lack of time. Two of them mentioned that asking for permission to participate in the study was a time-consuming procedure;
• Four indicated that they found recording conversations with clients inconvenient and inappropriate to the nature of the profession, with particular reference to the sanctuary function;
• Two made one (or more) attempts to record a conversation with a client without a desired result. Sometimes the audio recordings were not clearly intelligible, or no potential participant was present at the time;
• One indicated that he worked in the emergency room and the intensive care unit, and that it was undesirable, and often impossible, to record a conversation there;
• One indicated that he had only recently started working in the profession, and there was still some uncertainty about taking part in a study of this kind;
• Last, there was a humanist chaplain who, after repeated contact by telephone and e-mail, became unresponsive. The reason is unclear.

3.2. Audio Recordings

Research into the nature of verbal responses is inevitably linked to audio recording, as it is impractical to analyse verbal responses just by being present during a conversation. Conversations between humanist chaplains and clients are precarious in nature, making it undesirable to be present as a researcher. This is why the decision was made to work with audio recordings. We refrained from working with video recordings, although that is quite possible for conversation analysis (Have 2007). According to Have (2007, p. 8) audio recordings are sufficient and necessary for the analysis of verbal responses. Video recordings would help to analyse nonverbal responses, such as facial expressions and posture. This was not chosen for reasons of privacy since clients at a hospital are often in a room with several people.

In advance, the participants asked the clients for verbal permission to make an audio recording. After the conversation, the clients were asked to review the informed consent to participate in the study and to place a signature.

Table 1. Information on the audio recordings; n_f—empathic following response; n_u—empathic understanding response; d_a—allowed directive response; d_u—unallowed directive response; %d_u—percentage of unallowed directive responses.

Audio	Duration in Minutes	Verbal Responses by the Humanist Chaplain	nf	nu	da	du	% du
1	24.31	237	152	16	44	25	11
2	31.32	182	48	21	70	43	24
3	82.00	327	145	35	41	106	31
4	29.30	230	133	8	52	37	16
5	26.26	335	219	15	55	46	14
6	25.16	147	80	12	31	24	16
7	46.32	336	179	29	39	89	28
8	81.56	649	413	22	30	184	28
9	30.37	259	144	34	42	39	15
10	53.05	330	102	22	44	162	50
11	86.52	638	473	18	46	101	15
12	50.33	134	28	9	17	80	60
13	70.02	326	103	49	31	143	44

shows the percentages for the different types of verbal responses per audio recording.

3.3. Analysis

The deductive content analysis was performed on the basis of four codes: the empathic following response (n_f); the empathic understanding response (n_u); the allowed directive response (d_a); and the unallowed directive response (d_u), as shown in Figure 1. The transcript and audio recording of a conversation were used for the analysis at the same time, so it was not only possible to read what was said, but also to hear how something was said. Here, the four main types of verbal responses of the humanist chaplains are analysed on the basis of the conversation analysis, using the transcription conventions shown in

Table 2. Transcription conventions (Have 2007).

Sequencing
[	A single left bracket indicates the point of overlap between the speakers’ onsets.
]	A single right bracket indicates the point where an utterance, or part of an utterance, terminates vis-à-vis another.
=	Equal signs, one at the end of one line and one at the beginning of the next, indicate no “gap” between the two lines. This is often called latching.
Timed intervals
(0.0)	Numbers in parentheses indicate elapsed time in silence by tenth of seconds, so (7.1) is a pause of 7 and 1/10 of a second.
(.)	A dot in parentheses indicates a tiny “gap” within or between utterances.
Characteristics of speech production
word	Underlining indicates some form of stress, via pitch and/or amplitude; an alternative method is to print the stressed part in italics.
::	Colons indicate a prolongation of the immediately prior sound. Multiple colons indicate a more prolonged sound.
-	A dash indicates a cutoff.
.,??,	Punctuation marks are used to indicate characteristics of speech production, especially intonation; they do not refer to grammatical units; an alternative is an italicized question mark: ?
	A period indicates a stopping fall in tone.
	A comma indicates a continuing intonation, such as when you are reading items from a list.
?	A question mark indicates a rising intonation.
,?	The combined question mark/comma indicates a rise that is stronger than comma, but weaker than a question mark.
↑ ↓	Arrows indicate marked shifts into higher or lower pitches in the utterance part immediately following the arrow.
WORD	Upper case indicates louder sounds than the surrounding talk.
°	Utterances, or utterance parts bracketed by degree signs are quieter than the surrounding talk.
<>	Right/left carets bracketing an utterance or utterance part indicate speeding up.
∙hhh	A dot-prefixed row of hs indicates an inbreath. Without the dot, the hs indicate an outbreath.
w(h)or	A parenthesized h, or a row of hs within a word, indicate breathiness, as in laughter, crying, etc.
Transcriber’s doubts and comments
()	Empty parentheses indicate the transcriber’s inability to hear what was said. The length of the parenthesized space indicates the length of the untranscribed speech. In the speaker designation column, the empty parentheses indicate the inability to identify a speaker.
(word)	Parenthesized words are especially dubious hearings or speaker identifications.
(())	Double parentheses contain the transcriber’s descriptions, rather than, or in addition to, transcriptions.

(Have 2007). The verbal responses, especially those classified as “unallowed directive” (d_u), were also analysed in an “open” inductive manner, consisting of three steps, as described by Boeije (2014). The first step was the open coding of the verbal responses of the humanist chaplains. The codes were not included in the deductive analysis scheme beforehand, allowing for a more complete and nuanced analysis. The second step was the axial coding of the verbal responses, bringing the codes together and dividing them into several main codes. A code tree was drawn up on the basis of this distribution. The third and final step was the selective coding: ordering the codes on the basis of the problem definition. The most relevant codes for the underlying research were selected on the basis of this last step. On the basis of this phased analysis, the most relevant passages were selected and described in greater detail, in accordance with the transcription conventions as described by Have (2007) and included in Table 2.

4. Results

Figure 2 shows the average percentages of the different types of verbal responses in this study.

The results regarding the different types of verbal responses are discussed in detail in the following paragraphs.

4.1. Empathic Following Responses

On average, 49% of the verbal responses (range of 21–74%) constituted empathic following responses (n_f). When more empathic following responses were expressed, there was usually a conversation where the client mainly spoke to share his or her story. In this case, the humanist chaplain was there to “follow” the client, and empathic following responses were mainly expressed. When the humanist chaplain was “following” without expressing verbal responses, the percentage of empathic following responses (n_f) was clearly lower (Audio recording 12). This also applied to conversations where the humanist chaplain was more directively present (see Audio recording 10). Although the nature of most of the verbal responses was similar to the empathic following responses described and practised by Rogers, the ways in which they were expressed were sometimes different. The emphasis of the humanist chaplain with the client could more often be actively heard. The nuances were clear from the breathing of the humanist chaplain and his less monotonous way of speaking. As a result of the compassionate way of speaking, the empathic following responses sometimes exhibited a transition into verbal responses that were more “soothing” in nature. The humanist chaplain often automatically expressed these verbal responses in an effort to provide comfort to the client. Although these comforting responses were directively allowed (d_a), the humanist chaplain continued to follow the client unabatedly. The following passage is an example of this. A client’s lower leg had recently been amputated. The pain was much less than before, but he felt that his lower leg was still “present”:

Audio recording 2 (2.21–3.54)

C:	I feel:: the piece of leg that is still there. (1s) And: I feel what is no longer
	there, is my knee. (2s) A:nd I feel my hee:l bo:ne.
Hda:	°Ah:°
C:	What is no longer there: (2s) It does not hurt either, but I do feel them. They
	are here. Present. Whereas before that, especially my knee, it was just pain.
Hda:	°Tja:::°
C:	Only pain.
Hda:	°Bo↑y: oh boy°
C:	And now I feel it is there.
Hda:	°So::°

This conversation makes it clear that the humanist chaplain spoke in a soft, comforting way. Sometimes the humanist chaplain and the client spoke at the same time. This happened, in particular, when the humanist chaplain repeatedly said he was “following” the client. At that point, it seemed as though the humanist chaplain was interrupting the client, although it was not clear whether this was experienced in a negative way. The following conversation is an example of this:

Audio recording 5 (3.38–3.58)

C:	Yesterday I was examined.
Hnf:	Yes.
C:	By an assistant physician.
Hnf:	Y:es.
C:	This morning she came again
Hnf:	Huhu.
C:	and then she was going to discuss it all.
Hnf:	Yes:
C:	With her supervisor and
Hnf:	°Hmhm°
C:	then they would also go through the material from Cran Canaria,
Hnf:	Huhu.
C:	because she had deliberately not done that. I think that’s good too ·hhh
	because then at least you look with an open mind uh::
Hnf:	Yes:
C:	·hhh at to a patient.

This conversation makes it clear that the high frequency of empathic following responses shows that the client was constantly interrupted during his story.

4.2. Empathic Understanding Responses

On average, 8% of the verbal responses (range of 3–15%) were empathic understanding responses (n_u). Expressing an understanding response, the humanist chaplain took the floor, interrupting the client for a while. These verbal responses were expressed, for example, to check whether the humanist chaplain could still follow the client. In doing so, he presented the client, in a reflective way, with what he had understood. An empathic understanding response situates him more across from the client. We suspect that this played a part in the low average of 8%, probably mainly because the humanist chaplain wanted to be close to the client.

According to Rogers’ approach, an empathic understanding response should be used in a tentative manner, so that the client knows that he is allowed to respond to the words of the humanist chaplain and that there is no judgment being made. This kind of verbal response could be sporadically heard, though a nuance could be heard in how the empathic understanding response was expressed. In the following example the humanist chaplain made an effort to maintain harmony with the client, verbally responding in an even more cautious way:

Audio recording 13 (2.22–4.03)

C:	And then I ↑feel (2s) When I am just at home (2s) Alone or before with my hus:band, well then then you could be more yourself I will just say. You could also go ahea:d. You know, you could go out the doo:r and I cannot do that anymore. ·hhh <But> (1s) here I think, °then the next one comes in, and this was agreed°, but uh: (2s) yesterday a volunteer came here. ·hhh And everytime I think (2s) I know, I do not know how to ↑deal with that.
Hnf:	Y↑es.
C:	>So< restless.
Hnf:	[↑Hmhm. (2s) Yes:? Then you have the feeling of, there are ↑so many different disci↑plines here,
C:	>Yes<
Hnu:	and they all mean it ↑well. (1s) And somewhere I would like to ↑cooperate with them, ↑you know?
C:	Yes.
Hnf:	You ↑know?
C:	Sure.
Hnu:	Y↑es. You know? But it ↑costs me so much.
C:	·hhh It costs me so much, and sometimes I think ·hhh that it’s at the expense of myself.

The humanist chaplain’s high cautious questioning tone ensured that his understanding did not unexpectedly clash with the client’s point of view. The cautious tone was expressed, in particular, by the recurring phrase ‘’you know?’’

4.3. Directive Responses

On average, 16% of the verbal responses (range of 5–38%) were allowed directive (d_a). We refer to them as allowed verbal responses because, unless they are used sporadically, they do not stand in the way of a nondirective approach. The percentage of allowed directive responses of one humanist chaplain was considerably higher. This was because the humanist chaplain and the clients talked to each other as if it was a friendly conversation. The conversations were of a more everyday nature and they both spoke back and forth. These clients had been counselled by the humanist chaplain for a long time.

On average, 27% of the verbal responses (range of 11–60%) were “unallowed directive” in nature (d_u). We can, thus, note that the approach of the humanist chaplain was not entirely consistent with that of Rogers. The average percentage was higher in the twelfth audio recording, since few verbal following responses were expressed.

4.4. A Shift during the Conversation

The inductive analysis showed that, on four occasions, there was a certain “twist” (unallowed directive response), where the humanist chaplain tried to shift the focus of the conversation. He presumably felt a certain discomfort because of what the client said, focusing on his own inability and grief. At such a moment, the humanist chaplain expressed an unallowed directive response to allow the client to look at his life with hope again. Three out of four times, this kind of verbal response was used by the same participant. In the following conversation between the client and the humanist chaplain, the client shared that she had imagined life very differently. Her sadness was audible:

Audio recording 6 (14.50–15.50)

C:	I do not say this very often, but I imagined life very differently. ((audible
	sadness))
	(1s)
Hnf:	↑Hm↓hm.
C:	·hhh That I could help my son in the bu::siness with cleaning and stuff.
Hnf:	↑Hm↓hm.
C:	You know? Could clean his house. ·hhh Or do his laundry ·hhh
	(2s)
Hdu:	And if you look at your life? What are you most °happy about°?

It is clear that the humanist chaplain did not respond to the client’s disappointment. Instead, the humanist chaplain posed a question about what the client was most happy about in life. The humanist chaplain was redirecting the conversation to get away from the client’s grief.

4.5. Repeating Words

When the humanist chaplain seemed uncertain about what the client actually meant, he often repeated the client’s words in a slow, soft tone. Unlike a question, which would briefly situate him across from the client, the closeness seemed better preserved by this verbal response. When he repeated the client’s words, the client presented a clearer and more nuanced story. The client seemed called upon to explain a bit more, without being explicitly asked to do so.

During the conversation described below, the humanist chaplain asked how hope had returned to the client’s life. The client replied twice with a word (“internally” and “message”), which the humanist chaplain did not immediately understand. As he slowly repeated the client’s words, the client came up with a further explanation:

Audio recording 1 (11.15–12.09)

Hdu:	And how did the hope return?
C:	°Yes° (2s) inter::nally.
Hnf:	°Internally°.
	(2s)
C:	I suddenly got (.) A a just like you got a mes(hhh)sage.
	(1s)
Hnf:	A ↑°message°.
C:	↓Yes. (.) someone was with me all the time.
Hnf:	↑Hm↓hm.
C:	While no::body was here (.) I felt someone (.) who was with me.

The gentle, careful way the humanist chaplain repeated a word said by the client seemed inviting, as if the client was implicitly given permission to discuss it further. This ensured that the humanist chaplain received a more substantive explanation.

5. Discussion

In conversations with hospital clients, humanist chaplains tend to rely on an accordance with Rogers’ nondirective approach. On average, 57% of the verbal responses were nondirective, and 16% of the directive verbal responses were allowed. Despite this, 27% of the verbal responses, unlike Rogers’ approach, were unallowed directive. In addition, certain nuances could be heard in the verbal responses. It was clear, for example, that humanist chaplains were more compassionate and sometimes used verbal responses to comfort the client. Relatively few understanding responses were expressed, presumably because they could unintentionally interrupt the clients during their story. Note that there were differences in how empathic following responses and empathic understanding responses were expressed. Sometimes the humanist chaplain expressed empathic following responses very frequently, so that he and the client sometimes spoke at the same time. A humanist chaplain who hardly expressed empathic following responses and empathic understanding responses in a verbal way could use the nondirective approach. This is because the nonverbal responses could also act as empathic following responses.

The inductive analysis shows that the humanist chaplains sometimes repeated the client’s words if they felt they might not have understood the meaning of these words or if did not sufficiently understand the client. Rogers (1975) long struggled with the caricatural image of his nondirective approach, which involves reflective listening. Reflective listening, according to this caricatural view, merely involves the counsellor repeating the client’s last words. Although reflective listening is more than just repeating words, this study shows that repeating words can be inviting and helpful to the client. The repeating response by the humanist chaplain was, thus, an empathic following response and, at the same time, an empathic understanding response. If the client discussed his inability, grief, and pain, unallowed directive responses were sometimes expressed, possibly to divert the client’s attention from his own powerlessness. Humanist chaplains asked about the valuable and hopeful moments in the client’s life. We do not know how this kind of unallowed directive response, whereby the humanist chaplain did not dwell on the client’s expression of inability, grief, and pain, was experienced by the client and whether it had adverse consequences. However, according to Jorna (2008, p. 275), the chaplain is there to turn to the client in an open way and to be open to his “place of trouble”. He also noted that it could be difficult to devote attention to what is undesirable. The findings of the current study support this observation.

The conversational analysis was helpful in examining the verbal responses, since it revealed what the humanist chaplains were saying and how they said it. On the basis of the conversation analysis, it is clear that the verbal responses were sometimes expressed in a more comforting way. Because of their profession, the chaplains were likely to represent ’The Good’ (Schuhmann and Damen 2018) and could be expected to guide their clients to a consoling point of view. Despite the clear overlap between soothing responses and providing an accepting and comforting atmosphere in the context of client-centred (Rogerian) psychotherapy, from a linguistic point of view, some chaplains practiced unallowed directive responses. The researchers had the impression that this did not always match with their clients’ expectations. Here, our study offers a point of departure for further enquiry, such as how a Rogerian approach and interventional aspects of spiritual and moral guidance can be tolerated or integrated.

A limitation of the study is that the nonverbal responses were not taken into account. In addition, only humanist chaplains who worked in a general hospital setting in the Netherlands were approached. No other fields, denominations, or countries were included. Most of the humanist chaplains who were approached did not participate in the study. In his research on the (lack of) empirical research on hospital chaplaincy, Carrigan (1966, p. 47) concludes: “It is clear that in pastoral care research we have barely begun to scratch the surface”. According to Carrigan (1966), this is due to: (1) A lack of time; (2) A lack of permission from (the management of) the hospital; (3) An existing tension between religion and science; and (4) An uncertainty on the part of the chaplains about examining themselves. Points 1 and 4 were confirmed by some of the humanist chaplains in this study. Although the humanist chaplains did not talk about a tension between religion (or philosophy of life) and science, some did mention that, given the nature of their profession, it was inconvenient to record conversations. In this respect, we can cautiously state that Carrigan’s conclusion (1966) is still current. Another limitation is that there was only one researcher available to analyse and score the conversations. No interjudge agreement could be computed to determine the reliability of the scoring. However, the typology of the chaplains’ utterances was straightforward and evident, which made a second scoring less necessary.

An option for further research might be to analyse audio recordings on the basis of other approaches. This could make clearer what (elements of) other approaches were used during conversations. A second option might be to conduct a similar study among a larger population of humanist or other chaplains. In addition, we also recommend conducting research into the effects of verbal and, optionally, nonverbal responses. It could help to explicitly ask clients about the effects of verbal responses and conversations in general. The present study shows how humanist hospital chaplains offer a, by and large, Rogerian approach to their clients. However, their interventions tend to lean towards showing compassion, even up to the point of distraction, as well as to showing guidance towards issues of hope or human dignity. How this combination of a Rogerian approach and spiritual guidance should be evaluated is uncertain and deserves further investigation.

Since chaplains, in general, adopted Rogers’ nondirective approach around the mid-20th century (see, e.g., Rogers 1951; Stone 2001; O’Connor 2003; Snodgrass 2007; and Carey and Cohen 2015), the current study is relevant for chaplains of various denominations. Because of recent changes in the religious landscape of the Netherlands and other Western societies, chaplains increasingly meet people from various cultural and religious backgrounds (Liefbroer and Berghuijs 2019). As a result, generic approaches, such as Rogers’ nondirective approach, presumably provide a good foundation for chaplains in general when conducting a conversation with a client.