diff --git a/jedi/api/__init__.py b/jedi/api/__init__.py
index 915c89ea..8961fa35 100644
--- a/jedi/api/__init__.py
+++ b/jedi/api/__init__.py
@@ -27,7 +27,7 @@ from jedi.api import helpers
 from jedi.api.completion import Completion
 from jedi.evaluate import Evaluator
 from jedi.evaluate import imports
-from jedi.evaluate.param import try_iter_content
+from jedi.evaluate.arguments import try_iter_content
 from jedi.evaluate.helpers import get_module_names, evaluate_call_of_leaf
 from jedi.evaluate.sys_path import get_venv_path, dotted_path_in_sys_path
 from jedi.evaluate.filters import TreeNameDefinition
diff --git a/jedi/evaluate/analysis.py b/jedi/evaluate/analysis.py
index f5cc6d7b..c825e5fe 100644
--- a/jedi/evaluate/analysis.py
+++ b/jedi/evaluate/analysis.py
@@ -181,7 +181,7 @@ def _check_for_exception_catch(node_context, jedi_name, exception, payload=None)
             assert trailer.type == 'trailer'
             arglist = trailer.children[1]
             assert arglist.type == 'arglist'
-            from jedi.evaluate.param import TreeArguments
+            from jedi.evaluate.arguments import TreeArguments
             args = list(TreeArguments(node_context.evaluator, node_context, arglist).unpack())
             # Arguments should be very simple
             assert len(args) == 2
diff --git a/jedi/evaluate/arguments.py b/jedi/evaluate/arguments.py
new file mode 100644
index 00000000..02000ce5
--- /dev/null
+++ b/jedi/evaluate/arguments.py
@@ -0,0 +1,247 @@
+from parso.python import tree
+
+from jedi._compatibility import zip_longest
+from jedi import debug
+from jedi.evaluate import analysis
+from jedi.evaluate.lazy_context import LazyKnownContext, LazyKnownContexts, \
+    LazyTreeContext, get_merged_lazy_context
+from jedi.evaluate.filters import ParamName
+from jedi.evaluate.base_context import NO_CONTEXTS
+from jedi.evaluate.context import iterable
+from jedi.evaluate.param import get_params, ExecutedParam
+
+def try_iter_content(types, depth=0):
+    """Helper method for static analysis."""
+    if depth > 10:
+        # It's possible that a loop has references on itself (especially with
+        # CompiledObject). Therefore don't loop infinitely.
+        return
+
+    for typ in types:
+        try:
+            f = typ.py__iter__
+        except AttributeError:
+            pass
+        else:
+            for lazy_context in f():
+                try_iter_content(lazy_context.infer(), depth + 1)
+
+
+class AbstractArguments(object):
+    context = None
+
+    def eval_argument_clinic(self, parameters):
+        """Uses a list with argument clinic information (see PEP 436)."""
+        iterator = self.unpack()
+        for i, (name, optional, allow_kwargs) in enumerate(parameters):
+            key, argument = next(iterator, (None, None))
+            if key is not None:
+                raise NotImplementedError
+            if argument is None and not optional:
+                debug.warning('TypeError: %s expected at least %s arguments, got %s',
+                              name, len(parameters), i)
+                raise ValueError
+            values = NO_CONTEXTS if argument is None else argument.infer()
+
+            if not values and not optional:
+                # For the stdlib we always want values. If we don't get them,
+                # that's ok, maybe something is too hard to resolve, however,
+                # we will not proceed with the evaluation of that function.
+                debug.warning('argument_clinic "%s" not resolvable.', name)
+                raise ValueError
+            yield values
+
+    def eval_all(self, funcdef=None):
+        """
+        Evaluates all arguments as a support for static analysis
+        (normally Jedi).
+        """
+        for key, lazy_context in self.unpack():
+            types = lazy_context.infer()
+            try_iter_content(types)
+
+    def get_calling_nodes(self):
+        raise NotImplementedError
+
+    def unpack(self, funcdef=None):
+        raise NotImplementedError
+
+    def get_params(self, execution_context):
+        return get_params(execution_context, self)
+
+
+class AnonymousArguments(AbstractArguments):
+    def get_params(self, execution_context):
+        from jedi.evaluate.dynamic import search_params
+        return search_params(
+            execution_context.evaluator,
+            execution_context,
+            execution_context.tree_node
+        )
+
+
+class TreeArguments(AbstractArguments):
+    def __init__(self, evaluator, context, argument_node, trailer=None):
+        """
+        The argument_node is either a parser node or a list of evaluated
+        objects. Those evaluated objects may be lists of evaluated objects
+        themselves (one list for the first argument, one for the second, etc).
+
+        :param argument_node: May be an argument_node or a list of nodes.
+        """
+        self.argument_node = argument_node
+        self.context = context
+        self._evaluator = evaluator
+        self.trailer = trailer  # Can be None, e.g. in a class definition.
+
+    def _split(self):
+        if isinstance(self.argument_node, (tuple, list)):
+            for el in self.argument_node:
+                yield 0, el
+        else:
+            if not (self.argument_node.type == 'arglist' or (
+                    # in python 3.5 **arg is an argument, not arglist
+                    (self.argument_node.type == 'argument') and
+                     self.argument_node.children[0] in ('*', '**'))):
+                yield 0, self.argument_node
+                return
+
+            iterator = iter(self.argument_node.children)
+            for child in iterator:
+                if child == ',':
+                    continue
+                elif child in ('*', '**'):
+                    yield len(child.value), next(iterator)
+                elif child.type == 'argument' and \
+                        child.children[0] in ('*', '**'):
+                    assert len(child.children) == 2
+                    yield len(child.children[0].value), child.children[1]
+                else:
+                    yield 0, child
+
+    def unpack(self, funcdef=None):
+        named_args = []
+        for star_count, el in self._split():
+            if star_count == 1:
+                arrays = self.context.eval_node(el)
+                iterators = [_iterate_star_args(self.context, a, el, funcdef)
+                             for a in arrays]
+                iterators = list(iterators)
+                for values in list(zip_longest(*iterators)):
+                    # TODO zip_longest yields None, that means this would raise
+                    # an exception?
+                    yield None, get_merged_lazy_context(
+                        [v for v in values if v is not None]
+                    )
+            elif star_count == 2:
+                arrays = self._evaluator.eval_element(self.context, el)
+                for dct in arrays:
+                    for key, values in _star_star_dict(self.context, dct, el, funcdef):
+                        yield key, values
+            else:
+                if el.type == 'argument':
+                    c = el.children
+                    if len(c) == 3:  # Keyword argument.
+                        named_args.append((c[0].value, LazyTreeContext(self.context, c[2]),))
+                    else:  # Generator comprehension.
+                        # Include the brackets with the parent.
+                        comp = iterable.GeneratorComprehension(
+                            self._evaluator, self.context, self.argument_node.parent)
+                        yield None, LazyKnownContext(comp)
+                else:
+                    yield None, LazyTreeContext(self.context, el)
+
+        # Reordering var_args is necessary, because star args sometimes appear
+        # after named argument, but in the actual order it's prepended.
+        for named_arg in named_args:
+            yield named_arg
+
+    def as_tree_tuple_objects(self):
+        for star_count, argument in self._split():
+            if argument.type == 'argument':
+                argument, default = argument.children[::2]
+            else:
+                default = None
+            yield argument, default, star_count
+
+    def __repr__(self):
+        return '<%s: %s>' % (self.__class__.__name__, self.argument_node)
+
+    def get_calling_nodes(self):
+        from jedi.evaluate.dynamic import MergedExecutedParams
+        old_arguments_list = []
+        arguments = self
+
+        while arguments not in old_arguments_list:
+            if not isinstance(arguments, TreeArguments):
+                break
+
+            old_arguments_list.append(arguments)
+            for name, default, star_count in reversed(list(arguments.as_tree_tuple_objects())):
+                if not star_count or not isinstance(name, tree.Name):
+                    continue
+
+                names = self._evaluator.goto(arguments.context, name)
+                if len(names) != 1:
+                    break
+                if not isinstance(names[0], ParamName):
+                    break
+                param = names[0].get_param()
+                if isinstance(param, MergedExecutedParams):
+                    # For dynamic searches we don't even want to see errors.
+                    return []
+                if not isinstance(param, ExecutedParam):
+                    break
+                if param.var_args is None:
+                    break
+                arguments = param.var_args
+                break
+
+        return [arguments.argument_node or arguments.trailer]
+
+
+class ValuesArguments(AbstractArguments):
+    def __init__(self, values_list):
+        self._values_list = values_list
+
+    def unpack(self, funcdef=None):
+        for values in self._values_list:
+            yield None, LazyKnownContexts(values)
+
+    def get_calling_nodes(self):
+        return []
+
+    def __repr__(self):
+        return '<%s: %s>' % (self.__class__.__name__, self._values_list)
+
+
+def _iterate_star_args(context, array, input_node, funcdef=None):
+    try:
+        iter_ = array.py__iter__
+    except AttributeError:
+        if funcdef is not None:
+            # TODO this funcdef should not be needed.
+            m = "TypeError: %s() argument after * must be a sequence, not %s" \
+                % (funcdef.name.value, array)
+            analysis.add(context, 'type-error-star', input_node, message=m)
+    else:
+        for lazy_context in iter_():
+            yield lazy_context
+
+
+def _star_star_dict(context, array, input_node, funcdef):
+    from jedi.evaluate.context.instance import CompiledInstance
+    if isinstance(array, CompiledInstance) and array.name.string_name == 'dict':
+        # For now ignore this case. In the future add proper iterators and just
+        # make one call without crazy isinstance checks.
+        return {}
+    elif isinstance(array, iterable.AbstractIterable) and array.array_type == 'dict':
+        return array.exact_key_items()
+    else:
+        if funcdef is not None:
+            m = "TypeError: %s argument after ** must be a mapping, not %s" \
+                % (funcdef.name.value, array)
+            analysis.add(context, 'type-error-star-star', input_node, message=m)
+        return {}
+
+
diff --git a/jedi/evaluate/base_context.py b/jedi/evaluate/base_context.py
index 3af4f5b8..693a99aa 100644
--- a/jedi/evaluate/base_context.py
+++ b/jedi/evaluate/base_context.py
@@ -59,7 +59,7 @@ class Context(BaseContext):
         """
         Execute a function with already executed arguments.
         """
-        from jedi.evaluate.param import ValuesArguments
+        from jedi.evaluate.arguments import ValuesArguments
         arguments = ValuesArguments([ContextSet(value) for value in value_list])
         return self.execute(arguments)
 
diff --git a/jedi/evaluate/context/function.py b/jedi/evaluate/context/function.py
index aba8d9cd..0dba9c91 100644
--- a/jedi/evaluate/context/function.py
+++ b/jedi/evaluate/context/function.py
@@ -7,9 +7,9 @@ from jedi.evaluate import compiled
 from jedi.evaluate import recursion
 from jedi.evaluate import docstrings
 from jedi.evaluate import pep0484
-from jedi.evaluate import param
 from jedi.evaluate import flow_analysis
 from jedi.evaluate import helpers
+from jedi.evaluate.arguments import AnonymousArguments
 from jedi.evaluate.filters import ParserTreeFilter, FunctionExecutionFilter, \
     ContextName, AbstractNameDefinition, ParamName
 from jedi.evaluate.base_context import ContextualizedNode, NO_CONTEXTS, \
@@ -71,7 +71,7 @@ class FunctionContext(use_metaclass(CachedMetaClass, TreeContext)):
 
     def get_function_execution(self, arguments=None):
         if arguments is None:
-            arguments = param.AnonymousArguments()
+            arguments = AnonymousArguments()
 
         return FunctionExecutionContext(self.evaluator, self.parent_context, self, arguments)
 
diff --git a/jedi/evaluate/context/instance.py b/jedi/evaluate/context/instance.py
index 8e10981d..2c8d796c 100644
--- a/jedi/evaluate/context/instance.py
+++ b/jedi/evaluate/context/instance.py
@@ -8,7 +8,7 @@ from jedi.evaluate.base_context import Context, NO_CONTEXTS, ContextSet, \
     iterator_to_context_set
 from jedi.evaluate.lazy_context import LazyKnownContext, LazyKnownContexts
 from jedi.evaluate.cache import evaluator_method_cache
-from jedi.evaluate.param import AbstractArguments, AnonymousArguments
+from jedi.evaluate.arguments import AbstractArguments, AnonymousArguments
 from jedi.cache import memoize_method
 from jedi.evaluate.context.function import FunctionExecutionContext, FunctionContext
 from jedi.evaluate.context.klass import ClassContext, apply_py__get__
diff --git a/jedi/evaluate/context/iterable.py b/jedi/evaluate/context/iterable.py
index 7f725f7b..d0f468e4 100644
--- a/jedi/evaluate/context/iterable.py
+++ b/jedi/evaluate/context/iterable.py
@@ -534,7 +534,7 @@ def _check_array_additions(context, sequence):
     >>> a = [""]
     >>> a.append(1)
     """
-    from jedi.evaluate import param
+    from jedi.evaluate import arguments
 
     debug.dbg('Dynamic array search for %s' % sequence, color='MAGENTA')
     module_context = context.get_root_context()
@@ -543,7 +543,7 @@ def _check_array_additions(context, sequence):
         return ContextSet()
 
     def find_additions(context, arglist, add_name):
-        params = list(param.TreeArguments(context.evaluator, context, arglist).unpack())
+        params = list(arguments.TreeArguments(context.evaluator, context, arglist).unpack())
         result = set()
         if add_name in ['insert']:
             params = params[1:]
@@ -614,8 +614,8 @@ def get_dynamic_array_instance(instance):
         return instance.var_args
 
     ai = _ArrayInstance(instance)
-    from jedi.evaluate import param
-    return param.ValuesArguments([ContextSet(ai)])
+    from jedi.evaluate import arguments
+    return arguments.ValuesArguments([ContextSet(ai)])
 
 
 class _ArrayInstance(object):
@@ -643,8 +643,8 @@ class _ArrayInstance(object):
             for lazy in lazy_context.infer().iterate():
                 yield lazy
 
-        from jedi.evaluate import param
-        if isinstance(var_args, param.TreeArguments):
+        from jedi.evaluate import arguments
+        if isinstance(var_args, arguments.TreeArguments):
             additions = _check_array_additions(var_args.context, self.instance)
             for addition in additions:
                 yield addition
diff --git a/jedi/evaluate/context/klass.py b/jedi/evaluate/context/klass.py
index a040b33a..b7d61d3e 100644
--- a/jedi/evaluate/context/klass.py
+++ b/jedi/evaluate/context/klass.py
@@ -135,8 +135,8 @@ class ClassContext(use_metaclass(CachedMetaClass, TreeContext)):
     def py__bases__(self):
         arglist = self.tree_node.get_super_arglist()
         if arglist:
-            from jedi.evaluate import param
-            args = param.TreeArguments(self.evaluator, self, arglist)
+            from jedi.evaluate import arguments
+            args = arguments.TreeArguments(self.evaluator, self, arglist)
             return [value for key, value in args.unpack() if key is None]
         else:
             return [LazyKnownContext(compiled.create(self.evaluator, object))]
diff --git a/jedi/evaluate/dynamic.py b/jedi/evaluate/dynamic.py
index 01e18d61..7d05000d 100644
--- a/jedi/evaluate/dynamic.py
+++ b/jedi/evaluate/dynamic.py
@@ -22,7 +22,8 @@ from jedi import settings
 from jedi import debug
 from jedi.evaluate.cache import evaluator_function_cache
 from jedi.evaluate import imports
-from jedi.evaluate.param import TreeArguments, create_default_params
+from jedi.evaluate.arguments import TreeArguments
+from jedi.evaluate.param import create_default_params
 from jedi.evaluate.helpers import is_stdlib_path
 from jedi.evaluate.utils import to_list
 from jedi.parser_utils import get_parent_scope
diff --git a/jedi/evaluate/finder.py b/jedi/evaluate/finder.py
index efc36954..96032ae9 100644
--- a/jedi/evaluate/finder.py
+++ b/jedi/evaluate/finder.py
@@ -23,7 +23,7 @@ from jedi.evaluate.context import AbstractInstanceContext
 from jedi.evaluate import compiled
 from jedi.evaluate import analysis
 from jedi.evaluate import flow_analysis
-from jedi.evaluate import param
+from jedi.evaluate.arguments import TreeArguments
 from jedi.evaluate import helpers
 from jedi.evaluate.context import iterable
 from jedi.evaluate.filters import get_global_filters, TreeNameDefinition
@@ -231,7 +231,7 @@ def _check_isinstance_type(context, element, search_name):
 
         # arglist stuff
         arglist = trailer.children[1]
-        args = param.TreeArguments(context.evaluator, context, arglist, trailer)
+        args = TreeArguments(context.evaluator, context, arglist, trailer)
         param_list = list(args.unpack())
         # Disallow keyword arguments
         assert len(param_list) == 2
diff --git a/jedi/evaluate/param.py b/jedi/evaluate/param.py
index 72f598a7..a46394ce 100644
--- a/jedi/evaluate/param.py
+++ b/jedi/evaluate/param.py
@@ -1,20 +1,15 @@
 from collections import defaultdict
 
-from jedi._compatibility import zip_longest
-from jedi import debug
 from jedi.evaluate.utils import PushBackIterator
-from parso.python import tree
 from jedi.evaluate import analysis
-from jedi.evaluate.lazy_context import LazyKnownContext, LazyKnownContexts, \
-    LazyTreeContext, LazyUnknownContext, get_merged_lazy_context
+from jedi.evaluate.lazy_context import LazyKnownContext, \
+    LazyTreeContext, LazyUnknownContext
 from jedi.evaluate import docstrings
 from jedi.evaluate import pep0484
-from jedi.evaluate.filters import ParamName
-from jedi.evaluate.base_context import NO_CONTEXTS
 from jedi.evaluate.context import iterable
 
 
-def add_argument_issue(parent_context, error_name, lazy_context, message):
+def _add_argument_issue(parent_context, error_name, lazy_context, message):
     if isinstance(lazy_context, LazyTreeContext):
         node = lazy_context.data
         if node.parent.type == 'argument':
@@ -22,211 +17,6 @@ def add_argument_issue(parent_context, error_name, lazy_context, message):
         analysis.add(parent_context, error_name, node, message)
 
 
-def try_iter_content(types, depth=0):
-    """Helper method for static analysis."""
-    if depth > 10:
-        # It's possible that a loop has references on itself (especially with
-        # CompiledObject). Therefore don't loop infinitely.
-        return
-
-    for typ in types:
-        try:
-            f = typ.py__iter__
-        except AttributeError:
-            pass
-        else:
-            for lazy_context in f():
-                try_iter_content(lazy_context.infer(), depth + 1)
-
-
-class AbstractArguments():
-    context = None
-
-    def eval_argument_clinic(self, parameters):
-        """Uses a list with argument clinic information (see PEP 436)."""
-        iterator = self.unpack()
-        for i, (name, optional, allow_kwargs) in enumerate(parameters):
-            key, argument = next(iterator, (None, None))
-            if key is not None:
-                raise NotImplementedError
-            if argument is None and not optional:
-                debug.warning('TypeError: %s expected at least %s arguments, got %s',
-                              name, len(parameters), i)
-                raise ValueError
-            values = NO_CONTEXTS if argument is None else argument.infer()
-
-            if not values and not optional:
-                # For the stdlib we always want values. If we don't get them,
-                # that's ok, maybe something is too hard to resolve, however,
-                # we will not proceed with the evaluation of that function.
-                debug.warning('argument_clinic "%s" not resolvable.', name)
-                raise ValueError
-            yield values
-
-    def eval_all(self, funcdef=None):
-        """
-        Evaluates all arguments as a support for static analysis
-        (normally Jedi).
-        """
-        for key, lazy_context in self.unpack():
-            types = lazy_context.infer()
-            try_iter_content(types)
-
-    def get_calling_nodes(self):
-        raise NotImplementedError
-
-    def unpack(self, funcdef=None):
-        raise NotImplementedError
-
-    def get_params(self, execution_context):
-        return get_params(execution_context, self)
-
-
-class AnonymousArguments(AbstractArguments):
-    def get_params(self, execution_context):
-        from jedi.evaluate.dynamic import search_params
-        return search_params(
-            execution_context.evaluator,
-            execution_context,
-            execution_context.tree_node
-        )
-
-
-class TreeArguments(AbstractArguments):
-    def __init__(self, evaluator, context, argument_node, trailer=None):
-        """
-        The argument_node is either a parser node or a list of evaluated
-        objects. Those evaluated objects may be lists of evaluated objects
-        themselves (one list for the first argument, one for the second, etc).
-
-        :param argument_node: May be an argument_node or a list of nodes.
-        """
-        self.argument_node = argument_node
-        self.context = context
-        self._evaluator = evaluator
-        self.trailer = trailer  # Can be None, e.g. in a class definition.
-
-    def _split(self):
-        if isinstance(self.argument_node, (tuple, list)):
-            for el in self.argument_node:
-                yield 0, el
-        else:
-            if not (self.argument_node.type == 'arglist' or (
-                    # in python 3.5 **arg is an argument, not arglist
-                    (self.argument_node.type == 'argument') and
-                     self.argument_node.children[0] in ('*', '**'))):
-                yield 0, self.argument_node
-                return
-
-            iterator = iter(self.argument_node.children)
-            for child in iterator:
-                if child == ',':
-                    continue
-                elif child in ('*', '**'):
-                    yield len(child.value), next(iterator)
-                elif child.type == 'argument' and \
-                        child.children[0] in ('*', '**'):
-                    assert len(child.children) == 2
-                    yield len(child.children[0].value), child.children[1]
-                else:
-                    yield 0, child
-
-    def unpack(self, funcdef=None):
-        named_args = []
-        for star_count, el in self._split():
-            if star_count == 1:
-                arrays = self.context.eval_node(el)
-                iterators = [_iterate_star_args(self.context, a, el, funcdef)
-                             for a in arrays]
-                iterators = list(iterators)
-                for values in list(zip_longest(*iterators)):
-                    # TODO zip_longest yields None, that means this would raise
-                    # an exception?
-                    yield None, get_merged_lazy_context(
-                        [v for v in values if v is not None]
-                    )
-            elif star_count == 2:
-                arrays = self._evaluator.eval_element(self.context, el)
-                for dct in arrays:
-                    for key, values in _star_star_dict(self.context, dct, el, funcdef):
-                        yield key, values
-            else:
-                if el.type == 'argument':
-                    c = el.children
-                    if len(c) == 3:  # Keyword argument.
-                        named_args.append((c[0].value, LazyTreeContext(self.context, c[2]),))
-                    else:  # Generator comprehension.
-                        # Include the brackets with the parent.
-                        comp = iterable.GeneratorComprehension(
-                            self._evaluator, self.context, self.argument_node.parent)
-                        yield None, LazyKnownContext(comp)
-                else:
-                    yield None, LazyTreeContext(self.context, el)
-
-        # Reordering var_args is necessary, because star args sometimes appear
-        # after named argument, but in the actual order it's prepended.
-        for named_arg in named_args:
-            yield named_arg
-
-    def as_tree_tuple_objects(self):
-        for star_count, argument in self._split():
-            if argument.type == 'argument':
-                argument, default = argument.children[::2]
-            else:
-                default = None
-            yield argument, default, star_count
-
-    def __repr__(self):
-        return '<%s: %s>' % (self.__class__.__name__, self.argument_node)
-
-    def get_calling_nodes(self):
-        from jedi.evaluate.dynamic import MergedExecutedParams
-        old_arguments_list = []
-        arguments = self
-
-        while arguments not in old_arguments_list:
-            if not isinstance(arguments, TreeArguments):
-                break
-
-            old_arguments_list.append(arguments)
-            for name, default, star_count in reversed(list(arguments.as_tree_tuple_objects())):
-                if not star_count or not isinstance(name, tree.Name):
-                    continue
-
-                names = self._evaluator.goto(arguments.context, name)
-                if len(names) != 1:
-                    break
-                if not isinstance(names[0], ParamName):
-                    break
-                param = names[0].get_param()
-                if isinstance(param, MergedExecutedParams):
-                    # For dynamic searches we don't even want to see errors.
-                    return []
-                if not isinstance(param, ExecutedParam):
-                    break
-                if param.var_args is None:
-                    break
-                arguments = param.var_args
-                break
-
-        return [arguments.argument_node or arguments.trailer]
-
-
-class ValuesArguments(AbstractArguments):
-    def __init__(self, values_list):
-        self._values_list = values_list
-
-    def unpack(self, funcdef=None):
-        for values in self._values_list:
-            yield None, LazyKnownContexts(values)
-
-    def get_calling_nodes(self):
-        return []
-
-    def __repr__(self):
-        return '<%s: %s>' % (self.__class__.__name__, self._values_list)
-
-
 class ExecutedParam(object):
     """Fake a param and give it values."""
     def __init__(self, execution_context, param_node, lazy_context):
@@ -352,7 +142,7 @@ def get_params(execution_context, var_args):
     for key, lazy_context in non_matching_keys.items():
         m = "TypeError: %s() got an unexpected keyword argument '%s'." \
             % (funcdef.name, key)
-        add_argument_issue(
+        _add_argument_issue(
             parent_context,
             'type-error-keyword-argument',
             lazy_context,
@@ -367,40 +157,10 @@ def get_params(execution_context, var_args):
         first_key, lazy_context = remaining_arguments[0]
         if var_args.get_calling_nodes():
             # There might not be a valid calling node so check for that first.
-            add_argument_issue(parent_context, 'type-error-too-many-arguments', lazy_context, message=m)
+            _add_argument_issue(parent_context, 'type-error-too-many-arguments', lazy_context, message=m)
     return result_params
 
 
-def _iterate_star_args(context, array, input_node, funcdef=None):
-    try:
-        iter_ = array.py__iter__
-    except AttributeError:
-        if funcdef is not None:
-            # TODO this funcdef should not be needed.
-            m = "TypeError: %s() argument after * must be a sequence, not %s" \
-                % (funcdef.name.value, array)
-            analysis.add(context, 'type-error-star', input_node, message=m)
-    else:
-        for lazy_context in iter_():
-            yield lazy_context
-
-
-def _star_star_dict(context, array, input_node, funcdef):
-    from jedi.evaluate.context.instance import CompiledInstance
-    if isinstance(array, CompiledInstance) and array.name.string_name == 'dict':
-        # For now ignore this case. In the future add proper iterators and just
-        # make one call without crazy isinstance checks.
-        return {}
-    elif isinstance(array, iterable.AbstractIterable) and array.array_type == 'dict':
-        return array.exact_key_items()
-    else:
-        if funcdef is not None:
-            m = "TypeError: %s argument after ** must be a mapping, not %s" \
-                % (funcdef.name.value, array)
-            analysis.add(context, 'type-error-star-star', input_node, message=m)
-        return {}
-
-
 def _error_argument_count(funcdef, actual_count):
     params = funcdef.get_params()
     default_arguments = sum(1 for p in params if p.default or p.star_count)
diff --git a/jedi/evaluate/stdlib.py b/jedi/evaluate/stdlib.py
index 8e39fb58..259d6927 100644
--- a/jedi/evaluate/stdlib.py
+++ b/jedi/evaluate/stdlib.py
@@ -13,7 +13,7 @@ import collections
 import re
 
 from jedi import debug
-from jedi.evaluate import param
+from jedi.evaluate.arguments import ValuesArguments
 from jedi.evaluate import analysis
 from jedi.evaluate import compiled
 from jedi.evaluate.context.instance import InstanceFunctionExecution, \
@@ -208,7 +208,7 @@ def builtins_reversed(evaluator, sequences, obj, arguments):
     # would fail in certain cases like `reversed(x).__iter__` if we
     # just returned the result directly.
     seq = iterable.FakeSequence(evaluator, 'list', rev)
-    arguments = param.ValuesArguments([ContextSet(seq)])
+    arguments = ValuesArguments([ContextSet(seq)])
     return ContextSet(CompiledInstance(evaluator, evaluator.BUILTINS, obj, arguments))
 
 
diff --git a/jedi/evaluate/syntax_tree.py b/jedi/evaluate/syntax_tree.py
index ed1b4db8..1d847a49 100644
--- a/jedi/evaluate/syntax_tree.py
+++ b/jedi/evaluate/syntax_tree.py
@@ -16,7 +16,7 @@ from jedi.evaluate import recursion
 from jedi.evaluate import helpers
 from jedi.evaluate import analysis
 from jedi.evaluate import imports
-from jedi.evaluate import param
+from jedi.evaluate import arguments
 from jedi.evaluate.context import ClassContext, FunctionContext
 from jedi.evaluate.context import iterable
 from jedi.evaluate.context import TreeInstance, CompiledInstance
@@ -149,8 +149,8 @@ def eval_trailer(context, base_contexts, trailer):
             )
         else:
             assert trailer_op == '('
-            arguments = param.TreeArguments(context.evaluator, context, node, trailer)
-            return base_contexts.execute(arguments)
+            args = arguments.TreeArguments(context.evaluator, context, node, trailer)
+            return base_contexts.execute(args)
 
 
 def eval_atom(context, atom):
@@ -528,7 +528,7 @@ def _apply_decorators(context, node):
             debug.warning('decorator not found: %s on %s', dec, node)
             return initial
 
-        values = dec_values.execute(param.ValuesArguments([values]))
+        values = dec_values.execute(arguments.ValuesArguments([values]))
         if not len(values):
             debug.warning('not possible to resolve wrappers found %s', node)
             return initial