###########################################################################################
### R code for the paper "Manner of motion in Estonian: A descriptive account of speed" ###
###########################################################################################

# Authors: Piia Taremaa and Anetta Kopecka
# Journal: Studies in Language
# Status of the paper: accepted


################
### PACKAGES ###
################

library(dplyr)
library(ggplot2)
library(sjPlot)
library(ggpubr)
library(Rcmdr)


################
### DATASETS ###
################

### DATA 1 ###
verbspeed = read.delim("PSG41.txt", header=T, encoding = "UTF-8")
verbspeed <- verbspeed %>% mutate_if(is.character,as.factor)
str(verbspeed)
verbspeed$Verb <- factor(verbspeed$Verb, levels = verbspeed$Verb[order(verbspeed$VerbSpeed)]) 
verbspeed$VerbType = factor(verbspeed$VerbType, levels = c("MannerVerb", "SourceVerb", "GoalVerb"))

### DATA 2 ###
man = read.delim("TaremaaKopecka_speed_21.11.21.txt", sep = "\t", header = T, encoding = "UTF-8")
man <- man %>% mutate_if(is.character,as.factor)
man$VerbType = factor(man$VerbType, levels = c("MannerVerb", "SourceVerb", "GoalVerb"))
man$MannerBroad = factor(man$MannerBroad, levels = c("yes", "no"))
man$Speed = factor(man$Speed, levels = c("yes", "no"))
man$SlowOrFast = factor(man$SlowOrFast, levels = c("slow", "fast", "variable"))
man$VerbSpeedBinned <- with(man, binVariable(VerbSpeed, bins=3, method='natural', labels=c('SlowVerbs','MediumVerbs','FastVerbs'))) # clustering in Rcmdr
colnames(man)
head(man)
str(man)
nrow(man)


##############
### TABLES ###
##############

### Table 1. Manner modifiers of narrow scope
# Instrument.Vehicle [==Vehicle], Speed, BodyMovements [==BodyMovements+OtherNarrow], Trajectory, FundConcepts, PostureWithoutPossession [==Posture/Position], PhysicalCond
summary(man)
# The total number of manner modifiers of narrow scope
A = filter(man, MannerNarrow == "yes")
nrow(A) # 1447

### Table 2. Manner modifiers of medium scope
# Emotion, Appearance.FaceClothing. [==Appearance], Possession, Sound, QualityOfMotion [==Quality], Observability
summary(man)
# The total number of manner modifiers of narrow scope and medium scope
1447+1029 # 2476
# The total number of clauses containing manner modifiers of narrow scope and/or medium scope
B = filter(man, MannerNarrow == "yes" | MannerMedium == "yes")
nrow(B) # 2344

### Table 3. Manner modifiers of broad scope
# OtherClearManner [==Other], NumberOfMovers, EnvirCharact [==Environment/atmosphere], Suddenness
summary(man)
# The total number of manner modifiers of narrow, medium, and broad scope
1447+1029+477 # 2953
# The total number of clauses containing manner modifiers of narrow, medium, and/or broad scope (control)
C = filter(man, MannerNarrow == "yes" | MannerMedium == "yes" | MannerLessClear == "yes")
nrow(C) # 2780


###############
### FIGURES ###
###############

### Figure 2. Standardised mean speed ratings of the verbs. The smaller the mean, the slower the verb, and the larger the mean, the faster the verb
set_theme(geom.label.angle = 00, axis.textcolor.x = "black", 
          axis.textcolor.y = "black", geom.outline.color = "gray50", 
          geom.outline.size = 0.5, 
          geom.label.color = "black",
          axis.angle.x = 0, 
          axis.textcolor = "black", 
          base = theme_bw(),
          legend.pos = "bottom")
(verbplot <-ggplot(verbspeed, aes(x=VerbSpeed, y=Verb, fill=VerbType)) +
    geom_dotplot(binaxis='y', stackdir='center', dotsize=0.5) +
    scale_fill_manual(values=c("gray98", "gray70", "black")) +
    theme(legend.position=c(0.8, 0.1), legend.direction="vertical")) # dot plot


### Figure 3. A: formal realisations of speed modifiers. B: formal realisations of manner modifiers of broad scope (speed modifiers included)
(p8 = plot_frq(man, SpeedForm, coord.flip = T, sort.frq = "asc", show.na = F, geom.colors = "gray90",
               hjust = "inward", vjust = "middle"))
(p9 = plot_frq(man, MannerForm, coord.flip = T, sort.frq = "asc", show.na = F, geom.colors = "gray90",
               hjust = "inward", vjust = "middle"))
(figure <- ggarrange(p8, p9,
                     labels = c("A", "B"),
                     ncol = 2, nrow = 1))


### Figure 4. The morphosyntactic realisation of speed modifiers across slow, fast, and variable motion
man$SpeedForm = factor(man$SpeedForm, 
                       levels = c("AdpP", "Comparison", "NP_part",
                                  "Combination", "NP_ine", "Gerund",
                                  "QuantP", "NP_ade", "NP_com", "AdvP"))
man2 <- man %>% group_by(SpeedForm) %>%   mutate(count_name_occurr = n())
ggplot(data=subset(man2, !is.na(SpeedForm)), aes(x=SpeedForm, fill=SlowOrFast))+
  geom_bar(position="fill") +
  scale_y_continuous(labels = scales::percent) +
  scale_fill_manual(values=c("gray95", "gray65", "gray40")) +
  geom_text(aes(label=..count..), stat='count', position=position_fill(vjust=0.5)) +
  xlab("Morphosyntactic form of speed modifiers") +
  ylab("") +
  coord_flip()



### Figure 5. A: the presence (= yes) or absence (= no) of speed modifiers across verb types. B: the presence (= yes) or absence (= no) of manner modifiers (broad scope, speed modifiers included) across verb types.
(p1 = ggplot(man, aes(x=VerbType, fill=Speed))+
    geom_bar(position="fill")+
    scale_y_continuous(labels = scales::percent) +
    scale_fill_manual(values=c("gray95", "gray75")) +
    geom_text(aes(label=..count..), stat='count', position=position_fill(vjust=0.5)))
sjt.xtab(man$VerbType, man$Speed, show.row.prc = T) # statistics
(p2 = ggplot(man, aes(x=VerbType, fill=MannerBroad))+
    geom_bar(position="fill")+
    scale_y_continuous(labels = scales::percent) +
    scale_fill_manual(values=c("gray95", "gray75")) +
    geom_text(aes(label=..count..), stat='count', position=position_fill(vjust=0.5)))
sjt.xtab(man$VerbType, man$MannerBroad, show.row.prc = T) # statistics
(figure <- ggarrange(p1, p2,
                     labels = c("A", "B"),
                     ncol = 2, nrow = 1))


### Figure 6. A: speed ratings of the verbs across the clauses with the presence (= yes) or absence (= no) of speed modifiers. B: speed ratings of the verbs across the clauses with the presence (= yes) or absence (= no) of manner modifiers (broad scope, speed modifiers included). C: general frequencies of speed modifiers that express fast, slow, or variable speed of motion. D: speed ratings of the verbs across the clauses with speed modifiers that express fast, slow, or variable speed of motion. E: speed modifiers across the verbs grouped into slow, medium, and fast motion. 
(p3 = ggplot(man, aes(x=Speed, y=VerbSpeed), rev.order = T) +
  geom_boxplot(notch = T) +
  xlab("The presence of speed modifiers") +
  theme(axis.title.x = element_text(face="bold", color="black")))
(p4 = ggplot(man, aes(x=MannerBroad, y=VerbSpeed)) +
  geom_boxplot(notch = T) +
  xlab("The presence of manner modifiers") +
  theme(axis.title.x = element_text(face="bold", color="black")))
(p5 = ggplot(data=subset(man, !is.na(SlowOrFast)), aes(x = SlowOrFast)) +
  geom_bar(fill = "gray90", colour = "black") +
  xlab("Speed modifiers") + 
  ylab("Frequency") +
  geom_text(aes(label=paste0(sprintf("%1.1f", ..count../sum(..count..)*100),
                             "%\n", ..count..), y=0.5*..count..), stat = "count", vjust = 0.3, colour = "black") +
  theme(axis.title.x = element_text(face="bold", color="black")))
(p6 = ggplot(data=subset(man, !is.na(SlowOrFast)), aes(x=SlowOrFast, y=VerbSpeed)) +
  geom_boxplot(notch = T) +
  xlab("Speed modifiers") +
  theme(axis.title.x = element_text(face="bold", color="black")))
(p7 = ggplot(data=subset(man, !is.na(SlowOrFast)), aes(x=VerbSpeedBinned, fill=SlowOrFast))+
  geom_bar(position="fill")+
  scale_y_continuous(labels = scales::percent) +
  scale_fill_manual(values=c("gray95", "gray75", "gray50")) +
  xlab("Motion verbs") +
  geom_text(aes(label=..count..), stat='count', position=position_fill(vjust=0.5)) +  
  labs(fill = "Speed modifiers"))
sjt.xtab(man$SlowOrFast, man$VerbSpeedBinned, show.row.prc = T, show.col.prc = T) # statistics
(figure <- ggarrange(p3, p4, p5, p6, p7,
                     labels = c("A", "B", "C", "D", "E"),
                     ncol = 2, nrow = 3))